CDMP Building and Infrastructure Study

This report results from the investigations and research of a team of consultants to determine the processes used in the implementation of four projects that suffered damage from recent floods and hurricanes.

The main purpose of the study is to show where in the implementation process improvements can be made to mitigate or eliminate the frequent failures of infrastructure and building projects from extreme natural events.

It is considered that a report of this nature should not contain the names of the consultants or construction contractors engaged on the projects being studied. However the institutions involved in the development of the projects have been named as the processes used are in some detail unique to each institution. The findings of this study should therefore assist the institutions in developing revised procedures that will focus on disaster mitigation as well as on the economic and financial viability of the projects.

It must be recognized that the conclusions reached are relevant to the processes used for the implementation of the projects studied. The genesis of the problems found appears to be that, generally, there has not been an emphasis on securing facilities from the ravages of the environment in which they exist, and that the resources needed to do this effectively may not have been made known to the owners. In this connection the need for the employment of a review consultant has been recommended.

As the public sector projects examined were in general constructed more than 20 years ago, information on the processes used and on the problems encountered during implementation was difficult to find. Verbal reports were made to the consultants, but written documentation to substantiate all of the reports was not always available.

The focus of the report is therefore on the decision-making process and not on the engineering, physical planning or architecture of the projects studied. Most of the problems found can be solved if a cooperative effort in disaster mitigation is made by the financing agency and the owner or government concerned. The conclusions that can be drawn from the examination of each of the four projects are described in the separate reports of the case studies at Appendix A, B, C, and D. The summary conclusions are given in Section A of this report.

This report also includes a section on recommended implementation procedures for infrastructure and building projects. This section recommends a procedure for ensuring that the project designers build a hazard mitigation strategy into the project concept and ensure that owners and financing agencies are aware of the costs and benefits of hazard mitigation before construction is financed.

The process of the selection of a consultant and the work the consultant should do should be described in the Manual for Consulting Services. This manual should be submitted as a separate volume to facilitate discussion with financing agencies and with consultants. The manual contains information on:

a) The Dominica Port was constructed in 1974–1978 with financing from the Caribbean Development Bank (using USAID funds) and the Government of the Commonwealth of Dominica. Hurricane "David" in 1979 inflicted severe damage to the facility particularly to the reclamation and its buildings and led to the reconstruction of the protective armor and repair to other components of the facility.

b) The Norman Manley Law School and the Philip Sherlock Centre for the Creative Arts of the University of the West Indies in Jamaica, along with many University-owned houses, were damaged by hurricane "Gilbert" in 1989. Financing for the construction of the buildings was from the Government of Jamaica and from a private donation.

c) St. Lucia bridges were damaged by storms of 1994 and 1996. All six of the bridges damaged by the storms of 1994 and 1996 were examined. The two bridges that were examined in some detail are the Troumassee bridge, which was first constructed some 90 years ago and extensively rehabilitated in 1976, and the Caico bridge. The Troumassee bridge structure itself was not damaged, but one of the approaches was washed away. The Caico bridge was destroyed in the 1994 storm and rebuilt. The 1996 storm destroyed the rebuilt bridge.

d) The Grand Palazzo Hotel in St. Thomas, damaged by hurricane "Marilyn" in 1995, was studied as it is a private sector facility constructed under private sector business principles, and not in accordance with the procedures mandated by international financing agencies.

The study was carried out mainly by discussions with officials and consultants involved in the projects and by examining information in the project files. All persons interviewed provided as much information as was available. Visits were made to the project sites to gain an appreciation of the extent of the damage and reconstruction work carried out on each project.

c) The Ministry of Planning, the office of the National Disaster Coordinator and the Ministry of Communications and Works in St. Lucia.

The Caribbean Development Bank, the financing agency for the Dominica Port and for the Troumassee bridge in St. Lucia, allowed the consultant to examine the files which contained documentation on the appraisal of the projects and construction reports. This information gave insight into the implementation processes of infrastructure projects financed by the CDB, and the involvement of the governments and other donor agencies in the design and construction of the projects.

Discussions were held with the Deputy Director of the CDB, the entity responsible for infrastructure projects, to ascertain the role the bank played in monitoring projects after the financial commitment had been made. This role is considered to be vital in ensuring that the projects continue to perform adequately in keeping with its projected output. The Vice President of the CDB expressed interest in the results of the study, especially as it is intended to point out the areas where institutional borrowers should be intimately involved in the decisions affecting the integrity of the projects and in the monitoring of the implementation of the projects.

The cases studied also provide information which should be useful to the bank in its attempt to stiffen its operating policies to improve the borrowers’ involvement in vital decisions affecting hazard mitigation and consequently the life of the projects. Specifically, the report recommends that the use of appropriate building codes and standards, such as the Organization of Eastern Caribbean States (OECS) building codes and the Caribbean Uniform Building Code (CUBIC) for buildings and the development of appropriate infrastructure standards for projects financed by the bank such as roads and drainage, bridges, ports, airports, water supply and sewerage.

A visit was also made to Dominica by the associated consultant. The consulting engineering firm of Consulting Engineers Partnership Ltd. was involved in the construction of the port and in the reconstruction after hurricane "David". Approval for the examination of the project files was obtained by the Consulting Engineers Partnership.

The research and interviews with the persons involved in the implementation of the project showed that the project design concentrated on ensuring that the cost of the project can be sustained by revenues and/or an economic rate of return that is acceptable to the financing agency. This is normally based on the assumption that the facility as designed, constructed, and operated would conform to normal criteria for such facilities, that the revenues projected can be obtained, and that the output of the project will provide an economic benefit to the community.

The failure of the reclamation and buildings due to hurricane Gilbert led to a re-examination of the criteria for design of the facilities and for appraising the project.

Information on the construction and on the damage of the St. Thomas hotel was found in the files of a consultant who examined the project for another agency. The files gave a reasonably detailed account of the processes used for the design and management of construction of the hotel so that judgements can be made as to the effectiveness of the process for hazard mitigation. It was considered desirable to examine a private sector project to determine whether the implementation procedures of major private sector projects are significantly different from the procedures of the international financing agencies.

A visit was made to St. Lucia to discuss the problems of bridge failures with the Chief Engineer of the Ministry of Communications and Works. The National Disaster Coordinator arranged the meetings with the relevant government officials and provided information and photographs of the damaged bridges.

The chief engineer of the Ministry of Communication and Works in St. Lucia arranged for the consultant to travel around the island and to see the bridges that failed as a result of the storms in 1994 and 1996. This trip around the island revealed the extent of the damages incurred as a result of the two storms and the reconstruction work being done on some of the bridges destroyed. The Chief Engineer also discussed in depth the problems of design and maintenance of bridges and waterways.

The government of St. Lucia appointed a commission to study the efficiency of the Ministry of Communications and Works. Care had to be taken to ensure that the study being carried out was not seen to be a part of the government’s commission.

The Vice Chancellor of the University of the West Indies (UWI) at Mona Jamaica requested that the Estate Manager make information available concerning the project. The university officials interviewed provided all information available in their files. Discussions were also held with the former Estate Manager who was partly responsible for the reconstruction work after hurricane "Gilbert". The consultants for the Norman Manley Law School provided information on the design of the roof and the cause of failure.

The information in this report is based on the interviews held and the research that has been carried out.

CDB: Caribbean Development Bank
USAID United States Agency for International Development
CIDA Canadian International Development Agency
GOCD: Government of the Commonwealth of Dominica
GOSTL: Government of the St. Lucia
ODA: Overseas Development Administration
BDD: British Development Division
OECS Organization of Eastern Caribbean States
UWI: University of the West Indies
GOJ: Government of Jamaica
BAPE: Barbados Association of Professional Engineers
CCEO: Caribbean Conference of Engineers Association
HURDAT: National Hurricane Center Database
SEAOC : Structural Engineers Association of California
EC: Eastern Caribbean (Dollar)
NMLS: Norman Manley Law School
PSCCA: Phillip Sherlock Centre for Creative Arts
CUBIC: Caribbean Uniform Building Code
IBRD: International Bank for Reconstruction and Development
KSAC: Kingston and St. Andrew Corporation

This section provides summary results of the case studies of the four projects examined. The details of the studies and descriptions of the damages incurred are given in Appendices A, B, C and D. The implementation procedures used are discussed as it was found that the procedures affect the choice of consultants and materials for most of the projects studied.

The information required to answer most of the questions asked was not readily available, but discussions with consultants and administrators who were involved in the development of the original projects or who had examined the damaged projects provided answers to key questions.

The damage to the projects examined resulted from hurricanes and rainstorms. The extreme natural events were:

a) Hurricane "David" (August 1979) was a strong category 4 hurricane with sustained winds of over 160 miles per hour and gusts of 200 miles per hour. The high waves generated by this hurricane destroyed the reclamation of the Dominica deepwater port and the concrete bridges connecting the reclamation to the berthing platform.

b) Hurricane "Gilbert" (September 1988) was a category 2 or 3 hurricane with winds of 145 miles per hour. This hurricane damaged the roofs of the Norman Manley Law School and the Phillip Sherlock Centre for the Creative Arts in the University of the West Indies in Jamaica. The roofs of many other University buildings were also damaged.

c) Tropical storm "Debbie" produced rainfall of 15 inches on the night of September 10, 1994. It was estimated that in the interior of the island of St. Lucia, over 20 inches of rain fell during a four hour period from 2:00 am to 6:00 a.m. The tropical depression of October 26, 1996 produced about 2 inches of rainfall. The floods resulting from the heavy rainfall caused significant damage to the roads and many road bridges were destroyed or severely damaged.

d) Hurricane "Marilyn" was a category 2 hurricane. This hurricane damaged some of the buildings and ancillary facilities of the Grand Palazzo Hotel in St. Thomas in 1995.

a) The Dominica Deepwater Port, constructed in 1974 to 1978 and damaged by hurricane "David" in 1979.

b) The Norman Manley Law School of the University of the West Indies in Jamaica was damaged by hurricane "Gilbert". The Philip Sherlock Centre for the Creative Arts which was also damaged by "Gilbert", since the implementation process which led to the damage of this building was essentially the same as that of the Norman Manley Law School, the summary results do not contain further information on this facility.

c) Main road bridges in St. Lucia, the Troumassee bridge and the Caico bridge. Both bridges were damaged by the 1996 flood in St. Lucia. The Caico bridge was reconstructed after a 1994 storm. The hasty reconstruction of this bridge led to the destruction of the reconstructed bridge by the tropical depression of 1996.

d) The Grand Palazzo Hotel in St. Thomas. This facility was damaged by hurricane "Marilyn" in 1995.

Of the four projects studied, the implementation procedures of two of the projects—the Dominica Deepwater Port and the Troumassee bridge—were in accordance with the requirements of the Caribbean Development Bank. The CDB’s procurement procedures depend to a large extent on the source of funding. For instance, the CDB used USAID funds for its financing of the Dominica deepwater port. The USAID procurement guidelines had to be followed. This meant that procurement of goods and services had to be from eligible sources listed by that institution.

Similar procurement requirements govern the financing assistance from other agencies, such as BDD and CIDA, which were involved in financing engineering studies and a warehouse for the Dominica deepwater port. The procedures of these institutions require that before a commitment is made, the recipient country is to approve the choice of consultant, or supplier of hardware, generally from a short list of eligible firms. For the procurement of hardware the donor agencies (BDD and CIDA) require the supplier to submit to the recipient country drawings and specifications of the hardware to be supplied. Procurement of the hardware is subject to the approval of the recipient country. Hence the steel frame of the banana warehouse for the Dominica deepwater port and the Bailey bridges supplied to the government of St. Lucia as emergency assistance by the BDD were approved by the respective governments.

These general guidelines were followed in the procurement of consulting services for the preliminary study of the Dominica deepwater port and for the selection of consultants for the final design and inspection of construction of the facility. The consultant engaged by ODA to carry out the preliminary design was not eligible for selection of the detailed design, in accordance with the USAID procurement guidelines. In accordance with the general guidelines of the ODA, all studies and preliminary designs of the Dominica deepwater port were submitted to the GOCD for approval.

The principles guiding the CDB’s involvement with the financing of projects was enunciated by the CDB in 1972 as follows. For the scheme to be financed, it must be:

This was the principle guiding the appraisal of the project. There is no mention here of hazard mitigation.

The procurement of consulting services for the Norman Manley Law School and for the Grand Palazzo Hotel followed the standard practice for procurement of such services. In the case of the Law school, the consultants engaged for the final design had won the design competition for building and were considered to be knowledgeable and experienced in the design and construction of such buildings.

The procedure for the choice of the consultants for the Grand Palazzo Hotel was not documented, but the consultants chosen have proven records of competence in the design and construction of hotels. The financing arrangements for the project did not influence the choice of consultants or the procurement of construction materials.

5.1 General

The principal findings described in this sub-section are based on the review of the four projects. Recommendations to ensure that hazard mitigation is given priority in the design process are given in Section D.

5.2 Dominica Deepwater Port

a) Consultants were appointed by the ODA in 1969 to undertake the preliminary design of the deepwater port. The consultants were required by their Terms of Reference to pay due regard to the dangers of hurricanes, earthquakes, squalls and other weather conditions that may affect the port. The ODA also appointed the Delft Hydraulics Laboratory to carry out a study of the wave regime that would affect the site selected for the port. Unfortunately, the results of this study did not reach the consultants until the preliminary design had been completed.

b) The consultant, on reviewing the Delft study, confirmed that the final design would take into account the probability of hurricanes affecting the structure. The consultant also confirmed that the design of the land reclamation and its protective revetment would be based on the effect of the wave heights associated with squalls which the Delft study stated are frequent and which would lead to a maximum height of swells of 1.0 meter. The Delft study also indicated that damage due to hurricane waves has occurred rarely, which can be attributed to the limited depth in front of the coastline. Maximum wave heights can be expected of 1.5 meters at the most.

c) The Delft study also stated that the loss of use of the port due to heavy seas was estimated to be no more than 3 days per year. This loss of use was acceptable to the Government of Dominica. The final design of the port therefore concentrated on ensuring that the port would be financially and economically viable, if designed and constructed in accordance with the recommendations of the preliminary design.

d) The final design was therefore expected to take into account the effect of hurricane waves on the port structure and that the land reclamation would be designed for wave heights not exceeding 1.5 meters. The consultants confirmed that the wharf structure would be designed to accommodate the normal loadings of trucks and moveable equipment.

e) The decision of the consultants to design the facility to accommodate a 6 foot wave height and to design the land reclamation at +9 feet above mean sea level (AMSL) rather than at +15 feet, was consistent with their view that the port structure would be designed to withstand the forces generated by hurricanes, but that the other facilities need only be designed to withstand the wave heights associated with squalls.

f) There was also the major consideration of the capital costs of the facility. If the facility had been designed for greater wave heights, costs would have been increased beyond the capacity of the facility to pay for itself from the projected revenues. This aspect was one of the criteria for project financing by the CDB.

g) The design of the reconstructed facility after the hurricane damage incurred was discussed in depth with the consultants and the CDB staff and it was agreed that the facility should be designed to resist high seas generated by hurricanes. Efforts were taken to ensure that the revetment would be strong enough to prevent destruction of the reclaimed land by storms of the category of "David" and for this reason dolos were used in place of the boulders used originally. The economic analysis of the reconstruction therefore took into account the fact that without the reconstruction the port could not function.

h) The costs of construction and reconstruction of the facility after the damage caused by hurricane "David" are given in Table 3 of Section B. The reconstruction required to improve the resistance of the facility to hurricane forces cost about US $3.9 million in 1982, or about 40.7% of the original 1975 cost.

i) The principal benefits from the construction of this facility were based on the savings in cost from the more efficient shipment of bananas, the principal export of Dominica. Shipments of bananas resumed within one year after the port was damaged, as the berthing structure was not damaged and ships were able to moor alongside the wharf. Temporary repairs to the land reclamation allowed some passage of vehicles to the ship.

5.3 Norman Manley Law School

a) The damage to the Norman Manley Law School by "Gilbert" occurred because of inadequate fastening mechanisms for the roof planks. The cost of repair of this damage in 1990 was about US $28,800 (1975 dollars) or no more than 4.2% of the original cost of the facility. (See Table 3 of Section B). While this cost is relatively small, the estimated expenditures that would have been required at the time of construction to avoid the damage was estimated to be no more than US $13,000.

b) Information on the design and construction is available to allow tracking of the implementation process. The building was designed subsequent to a design competition, which required the consultants to provide conceptual drawings and information on the accommodation aspects of the project.

c) The consultants selected were experienced in the design of buildings in Jamaica, and used the appropriate building codes for the design. The conceptual design was extensively reviewed by the university and the government of Jamaica, the financing agency. The designers stated that the project design was in accordance with the standards for wind design of the British Standards Institution as used in Jamaica at that time.

d) The contract for the consultants was in accordance with the standard UK conditions for engineers and architects. It was expected that resistance to hurricanes and earthquakes would have been a design criterion as required by Design Stage I of the UK Association of Consulting Engineers. The final designs were also reviewed by the University and by the Government of Jamaica. There was however no specific mention of a hazard mitigation strategy in the consultants' drawings and specifications.

e) The roof failure was due to the lack of attention to the roof fixing details. The roof structure did not fail and there was no significant damage to the building structure or its furniture and fittings.

f) The university did not employ a manager or an independent engineer for this project as the management of all university construction was carried out by a standing committee of the university. Hence there was no independent technical review of the drawings and specifications for the roof and its fixings.

5.4 St. Lucia bridges

a) About six road bridges were completely destroyed in 1994 by floods caused by heavy rainfall (tropical storm "Debbie"). In 1996 there was a tropical depression in which about 8" fell on October 26, and 2" on November 21 [1]. This depression caused the damage to the Troumassee and the Caico bridges.

b) The St. Lucia bridges that failed were, except for the Troumassee, designed and constructed by direct labor (force account) by the Public Works Department generally more than thirty years ago. Design information was not precise, as the process of construction by direct labor does not require the level of information needed for construction by private contractor. The failures, as reported by the chief engineer of the Ministry of Works, were due to a number of causes, most of which resulted from greater run-off generated by the increased development in the watersheds, lack of maintenance of the structures and waterways, and scouring of the footings or approaches.

c) The north approach to the Troumassee bridge in St. Lucia was breached by the flood waters of the 1996 storm. Following the 1994 storm, a decision was taken to reduce the size of the waterway of the bridge so as to induce faster flows and inhibit accretion. This reduction in the size of the waterway was the primary cause of the damage to the riverbanks and to the north approach.

d) The Caico bridge failure is an example of hasty construction without regard for hazard mitigation. The old bridge, probably constructed more than thirty years ago, was destroyed by the 1994 floodwaters. A Bailey bridge was constructed on the remaining abutments without the construction of apron protection, wing walls or riverbank protection. The 1996 floodwaters overturned the north abutment through scouring of the footings of the abutments and erosion of the banks and hence destroying the bridge.

e) It is clear that wing walls, protection of the abutments, and riverbank protection should have been constructed immediately after the Bailey bridge structure had been erected. This would have prevented the failure of the Bailey bridge and erosion of the footings of the abutments and the riverbanks and the subsequent destruction of the bridge structure.

f) The reported rainfall of the 1994 storm in St. Lucia was variously calculated to be greater than a one-in-a-hundred-year storm. This may be so, but the main road bridges constructed in 1975 with financing assistance by CDB were not damaged, although according to the criteria given in the pre-investment study the bridges were designed to accommodate a one- in-25-years storm.

g) This pre-investment study contained information on the rainfall intensities and frequency for the various bridge sites. The government should have examined this study before deciding on the reduction of the waterway of the Troumassee bridge in 1995.

5.5 Grand Palazzo Hotel

b) The design consultants and contractors were experienced in the design and construction of similar buildings in St. Thomas. The management system employed for the design and construction of the hotel showed that the owner was very involved in the decisions, respecting the progress of the project and the procurement of all material required for a successful project.

c) The Government of St. Thomas has established a regulatory body with the responsibility of ensuring that the design and construction are in accordance with the building permits issued by the regulatory authority, and in accordance with the building codes approved by the government. The project was designed in accordance with the structural requirements of the Uniform Building Code (UBC).

d) While the management system ensured that the owner and other decision makers were aware of the progress of the construction, and were assured that the design and construction had been approved by the regulatory body in St. Thomas, there was no mention in the system of a strategy for hazard mitigation. The owner was not involved in decisions affecting the resistance of the hotel to the natural hazards.

e) The cause of the failures was not due primarily to the use of inappropriate building codes or standards, but due to the lack of attention to the hazard mitigation requirements of non-structural components such as the glass doors and windows, and to the inadequate design of the roof structure of one of the buildings. The fabric roof over the dining terrace could have been made less vulnerable to high winds. It might however have required some change to the architecture of the roof and this might not have conformed to the required architecture or aesthetics of the terrace.

f) The cost of reconstruction of the hotel after the hurricane was over $5 million. This cost could have been avoided if the simple design procedures had been effected.

Reconstruction of infrastructure facilities after major damage by hurricanes, floods or earthquakes, takes place under an atmosphere of urgency. Often, this leads to hasty design decisions, which do not always correct the problems that caused the failure in the first place. Experience in the examination of many facilities has shown that the cost of repair is typically much larger than the extra cost which would have been needed at the time of construction of the facilities to avoid the damage. The need to invest in mitigation measures is one that should be discussed explicitly between the owner and the designers so that an informed decision can be made on the level and cost of protection required. Obviously, this discussion and consequent decision should take place at the earliest stage in the project cycle.

This section examines the costs of construction and reconstruction of all four facilities studied and shows the increased costs that would have been required to mitigate or avoid the damage which occurred. The section also shows the relationship between the costs of reconstruction and the original construction cost, using appropriate deflators.

This facility was damaged in 1979 by hurricane "David", a strong category 4 hurricane [3]. Information on the design and construction of the Dominica Port gathered from the research of the files of the Caribbean Development Bank (the principal financing agency of the project) showed that the port structure—berthing platform and buildings—had been designed to withstand a category 3 hurricane.

The damage inflicted was mainly to the reclamation area and its buildings and to the approach trestles. The reclamation’s protective armor was almost completely destroyed. The berthing platform itself was not damaged.

The design of the reconstruction recognized the need to strengthen the protective armor of the reclamation to withstand the forces of waves generated by hurricanes of category 4. An expenditure of about 3.9 million US dollars was required to rehabilitate the facilities and to improve the protection to the reclamation area.

Hurricane "Gilbert" caused significant damage to the University buildings including the Norman Manley Law School. The roof slabs (3" thick Tectum Roof Deck planks) and waterproof covering (¾" mastic asphalt) were lifted off by the high winds with subsequent damage to the building interior.

The main cause of the roof failure was due to the inadequate fixings of the roof slab, exacerbated by the failure of the glass clerestory window. While the roof structure itself was not damaged, lack of attention to the non-structural elements—the roof coverings—led to the failure. The correction needed was minor, but an expenditure of some US $90,000 (or about US $28,800 in 1978 dollars) was required to rehabilitate the building. It is reported that the University took the opportunity while rehabilitating the building to carry out some deferred maintenance work, hence the recorded hurricane reconstruction costs may be overstated.

The analysis given in Table 3 looks at the financial implications of reconstruction of Troumassee bridge. After a 1994 flood, a decision was made to reduce the apertures of the bridge to induce increased flow and reduce silting. The failure of the north approach to the Troumassee bridge in 1996 was due to this change in the bridge apertures, as the succeeding flood destroyed the bridge approach. The corrective work required after the damage incurred by the 1996 rainfall was mainly to remove the barrier to the closed aperture and to reinstate the north approach to provide for vehicular traffic.

The longer term reconstruction work planned involves reconstruction of the north approach, additional protection to the approaches, re-channeling of the river to follow the natural flow and relocation of the flood bund to ensure discharge downstream of the bridge.

The description and failure mode of another bridge that was damaged by the flood waters, the Caico bridge in Millett, are given in Appendix C. The costs attendant upon the reconstruction of this bridge have not been analyzed.

The Ministry of Communications and Works has provided all construction costs and estimates.

This hotel was damaged by hurricane "Marilyn" in 1995. Design and construction information on the hotel in St. Thomas has been examined. The management process used for the design and construction was defined in detail by the owners.

The detailed costs of reconstruction of this private project are not available, but the total costs of reconstruction were found.

The increased costs in engineering and construction required to prevent the damage to the roof of the reception building and other collateral damage is given in Table 3. It is probable that the owners may not wish to change the roofing material of the dining terrace as the use of the sail cloth roofing is principally for aesthetic reasons. No costs for improving the sail cloth roofs have been estimated.

The original construction cost of the facility, including furniture and equipment, was US $28 million. The estimated cost of reconstruction of the damaged facilities would not be more that 20% of the original construction cost, or US $5.6 million. The estimate of additional mitigation cost is less than 0.1% of the original cost. Since the design of a hotel or tourist facility very often has more to do with the creation of an ambience than the need for safe construction of all of its components, some damages may be unavoidable.

Original project cost is the total cost of the original project as recorded by the owner or financing agency.

Cost of reconstruction is the cost incurred to repair/rehabilitate the facility to its original state and functioning.

Additional mitigation cost is the estimated cost of studies, engineering/architectural design and construction services in addition to those included in the original project, that would have been needed to avoid or significantly reduce the damage suffered.

Table 2 shows the financial implications of the decisions taken for the design and construction of the Dominica Port, the Norman Manley Law School, the Troumassee bridge and the Grand Palazzo Hotel. The costs do not include economic losses from the non-use of the port or of the other projects studied. One of the primary reasons for constructing the Dominica Port was to ensure the safe shipment of bananas. Banana production was destroyed by the hurricane, but shipments resumed in 1980 as the port was partially in use during the reconstruction period. The economic losses from the non-use of the Grand Palazzo Hotel may be significant. Similarly, if the Caico bridge in St. Lucia is not restored soon, there will be losses from the extra cost in transporting bananas to the Port in Castries. The north approach to the Troumassee bridge was temporarily repaired immediately after the damage was incurred, as this bridge is on the main route from Castries to Vieux Fort.

a) Failures were in large part preventable.
b) Owners were not made aware of the risks associated with the design concepts.
c) Assumptions of the consultants involved in the reconstruction efforts were that the design measures taken would mitigate damage from similar hurricanes.

a) The increase in project costs which would have been necessary to mitigate the ensuing damage.
b) The real cost of the reconstruction after the hurricanes.

The retrospective look at the reconstruction costs required the use of appropriate construction price indices. This information is not readily available in the statistical reports from Jamaica or Dominica, and the use of the consumer price indices gives false pictures as there were significant changes in the currency exchange rates in Jamaica vis a vis the US and UK currencies. Consequently, construction price deflators for non-residential building construction in Barbados from information supplied by Consulting Engineers Partnership were used in this analysis.

In the reconstruction of the Dominica Deepwater Port, the CDB has calculated that local costs accounted for about 16% of the reconstruction costs (Table A-1 in Appendix A). Building prices in Dominica (and to a large extent in Jamaica) are therefore significantly affected by the movement of material and prices in the countries from which the material is imported, that is generally the United States, Canada and the United Kingdom. Labor productivity in Dominica and Jamaica would not differ significantly from that in Barbados for a project such as a port or a building of the type of the Norman Manley Law School. It is therefore considered that the construction price deflators for Barbados can reasonably be used to deflate construction prices in Jamaica. Construction price deflators are given in Table 2.

Year	Price Indices	Year	Price Indices
1975	58.4	1990	183.5
1981	122.5	1997	210.0
1982	129.2

a)    If the the height of the reclaimed land had been raised to heights indicated in the study by the Delft Hydraulics Laboratory, the increased costs to the facility would have been approximately US $655,000. This is equal to 11.5% of the final construction costs.
b)    The damage cost to the port in 1982 was US $3,933,000 as estimated by the CDB in their appraisal report. This cost when deflated to the year 1975 (about mid period of construction of the facility) is about 41% of the original cost of the facility.
c)    The estimated increased cost required to avoid the damage is no more than 28% of the cost of reconstruction.
d)    On the basis of this analysis, the increased cost of 11.5% needed to avoid the damage from hurricane David should have been incurred in the original design and construction.

a) The repair cost in 1990 was US $90,000.
b) This cost when deflated to 1975, the year of construction of the facility, is equal to 4.2% of the original construction cost of the facility.
c) The estimated cost needed to avoid the damage was no more that about US $13,000, mainly for extra roof fixings and some design detailing.
d) This cost is 1.9% of the cost of the facility.
e) The small expenditure of US $13,000 could easily have been justified at the time of construction of the facility.
f) Experience with the similar administrative reviews indicates that it is normal that maximum attention be paid to the accommodation requirements and not on the requirements for hazard mitigation. It is presumed that the engineers employed will ensure that the building as designed is capable of resisting the known hazards. Procedures of the university should include the specific requirement that the consultants employed for the design of buildings provide detailed information on the hazard mitigation strategy being used.

a) The cost of reconstruction work required to stabilize the banks of the river at this bridge site is estimated to be US $120,000.
b) This cost when deflated to 1975, the year of construction of the present bridge, is 17.4% of the 1975 construction cost.
c) The estimated costs of additional engineering and construction that would have been required in 1975 to prevent the damage which occurred in 1996 is US $20,000 or 10.8% of the original construction cost.
d) The designers of the project in 1974 had made provisions for protection of the riverbanks. The immediate cause of the damage from the heavy rainfall in 1996 was the due to the reduction of the waterway in 1995. If this construction had not been undertaken, there may have been less damage to the north bridge approach. The riverbank protection now recommended by the Ministry of Communications and Works would have prevented the damage.

a) The reconstruction cost was estimated to be 5.6 million US dollars. As the damage to the hotel was incurred only 4 years after the hotel was constructed the cost of reconstruction when deflated to the year of construction is 5.3 million US dollars.
b) The damage could have been avoided if the construction of the roof that failed and fixings to the doors and windows had been improved. The estimated cost of ensuring that the hotel was resistant to the high winds would have been US $27,000, not including the cost of improving the quality of the sail roof of the dining terrace.

TABLE 3 - COSTS OF CONSTRUCTION AND RECONSTRUCTION OF SELECTED PROJECTS - (US DOLLARS)

Most infrastructure projects in the Commonwealth Caribbean are financed partly by an international financing agency and by the government concerned. The procedure given in this paper is intended to assist the governments, as owners of the projects, in arranging the necessary steps that should be taken for the implementation of a successful project. The employment of consultants for the design of the project requires knowledge of the work the consultant is expected to perform and how the work should be performed.

Table 4 describes the main functions of the consultants employed by the owners and the responsibilities of the consultants and owners. The involvement of the financing agency in the process is also described. It should be noted however that the advice given by the financing agency during the pre-investment study or any other preliminary stage does not commit the financing agency to the financing of the project. This commitment is made after an appraisal of the project and approval by the financing agency of the amount of the loan to be made and the conditions under which the loan will be made. The early involvement of the financing agency is considered an important step to ensure a successful project and due attention to hazard mitigation.

The use of a special consultant—a review or check consultant—is described. The building codes developed for the OECS include the requirement that a special inspector be employed for reviewing the plans and construction activities of major projects as defined in the codes. The use of such consultants is a mandatory requirement in French construction law, British Columbia, and other states for projects described in the appropriate code or law. In this paper the term "review consultant" is used, as it is descriptive of the responsibility of the special consultant.

The need for such a consultant is to ensure, with his advice to the owner, that the design and construction activities are carried out in accordance with the accepted technical standards and, for special projects, in accordance with the appropriate technology. The review consultant reports directly to the owner and to the regulating body on the acceptability of the project as designed and constructed.

The steps in the implementation chain of infrastructure projects to be taken by the consultants are (or should be):

The terms Design Stage I and Design Stage II, the descriptions of which follow, have been developed by the Association of Consulting Engineers of the UK to describe the work and obligations of the consulting engineer. Most engineers of the Commonwealth Caribbean understand these terms and their contracts for engineering design work are based on the work as described by the Association of Consulting Engineers.

The requirements of Design Stage I and Design Stage II do not specifically include hazard mitigation measures, but the consulting engineer is required to advise the client on the necessity for investigations which in his opinion are needed for an appropriate design. In Design Stage II, the consultant will be expected to prepare designs and tender documents based on the information gathered in Design Stage I.

The principles of hazard resistant design must be incorporated in the project development at the earliest possible stage so that the client and the financial agency can take appropriate decisions on hazard mitigation measures. Subsequent design and material choices made in the detailed design and construction will be based on these decisions.

The pre-investment study is a necessary preliminary to the decision to invest, and enables the owner to obtain all the relevant information needed for a loan application.

This study includes engineering work normally described in Design Stage I (including the conceptual design or in some cases architectural sketches and perspectives) plus the work required to determine the economic and financial feasibility of the proposed development. This would include estimates of the cost of the project, maintenance costs, operating costs, revenues, and/or savings from the project.

The conceptual design and preliminary costs would be based on the information provided in this study, and an appreciation of the impact of the development on the environment. The design pulls together all of the information available and researched and provides a sensible view of the solutions to the problems posed by the accommodation requirements, land use, demography, topography, and other physical and social characteristics of the development.

At this stage, therefore, decisions must be made about all aspects of the development which would be impacted by the natural hazards and which will have a negative impact on the environment. The basic cost of the development is generally dependent on the conceptual design and other decisions made in the pre-investment study. Consequently, the feasibility of the proposals is dependent on the information gathered in the pre-investment study.

The details of Design Stage II are based on the conceptual design and on the principles agreed upon in the pre-investment study. This stage provides the detail drawings and material types and quantities needed for a construction contract, and requires that all major decisions affecting the siting or design of the development will have been made.

The requirements of Design Stage I as given by the Association of Consulting Engineers include:

a) Investigating data and information relevant to the works which are reasonably accessible to the consulting engineer and considering any reports relating to the works which have either been prepared by the consulting engineer or else prepared by others and made available to the consulting engineer by the client.

b) Making any normal topographic survey of the proposed site of the works that may be necessary to supplement the topographical information already available to the consulting engineer.

c) Advising the client on the need to carry out any geotechnical investigations which may be necessary to supplement the geotechnical information already available to the consulting engineer, arranging for such investigations when authorized by the client, certifying the amounts of any payments to be made by the client to the persons or firms carrying out such investigations under the consulting engineer’s direction, and advising the client on the results of such investigations.

d) Advising the client on the need for arrangements to be made for carrying out special surveys, special investigations or model tests, and advising the client of the results of any such surveys, investigations or tests.

e) Consulting any architect appointed by the client in connection with the architectural treatment of the works.

f) Preparing such documents as are reasonably necessary to enable the client to consider the consulting engineer’s general proposals for the construction of the works in the light of the investigations carried out at this stage; and to enable the client to apply for approval in principle of the execution of the works in accordance with such proposals.

g) Geotechnical investigations, to determine foundation characteristics, and other specialist investigations, such as oceanography and beach movements, are usually carried out at this stage as advised by the consultant. Such information may be required to allow the consultant to site the facility advantageously taking into account the need for hazard resistance and environmental control.

a) Preparing designs and tender drawings in connection with the works.
b) Advising as to the appropriate conditions to be incorporated in any contract to be made between the client and the contractor.
c) Preparing such specifications, schedules and bills of quantities as may be necessary to enable the client to obtain tenders or otherwise award a contract for carrying out the works.
d) Advising the client as to the suitability for carrying out the works of persons and firms tendering and as to the relative merits of tenders, prices and estimates received for carrying out the works.

Reports by the consultants on the progress of the project are necessary to allow the owner and the financing agency to monitor the project. These reports would normally be submitted monthly at the same time as the consultant or building contractor submits the interim certificates for payments. The reports should contain information on:

The final report is very often neglected, but this report should be submitted by the consultants and approved by both the owners and the financing agency as a pre-requisite to the claim for final payment. The final report must contain all of the information listed above plus as-built drawings showing all construction details and variations made.

The examination of the projects that were damaged by hurricanes and floods showed that the failures were in large part preventable. The following measures should have been taken to avoid the damage:

a) The failures could have been prevented if the components or assemblies of the projects were constructed to resist the known hazards. Most of the problems found can be solved if a cooperative effort in disaster mitigation is made by the financing agency and the owner or government concerned early in the design process.

b) Hazard mitigation was not a specific determinant in the design of the building projects, although the building codes normally used by structural engineers do take into account the need to resist or accommodate all loads, including wind and earthquake loads. The infrastructure projects studied were designed in accordance with standard practice for such projects and are based on the determination of the loads to be borne by the project. The preliminary studies of the Dominica Port and the pre-investment study for the St. Lucia bridges determined the loads—wind and rainfall run-off, as well as the traffic and berthing loads—that the projects were expected to bear.

c) The CDB appraisal reports of the Dominica Deepwater Port and of the St. Lucia bridges calculated the financial and economic benefits of the projects being appraised, but did not specifically consider the strategy for hazard resistance or the costs and benefits of hazard mitigation. The strategy for resisting the known hazards and the consequential costs and benefits should have been considered when calculating the economic and financial benefits of the port construction.

d) The suppliers of the steel framed warehouse for the port should have been required to certify that the building was designed in accordance with the requirements of the Barbados Association of Professional Engineers (BAPE) wind code or similar code acceptable to the Government of the Commonwealth of Dominica (GOCD). It is not known what standards were used for the design of the banana warehouse, but the failure of this building is testimony to the inadequate design of the structure. However, the other warehouses designed to resist the forces of a category 3 hurricane also failed.

e) The design and construction of the Norman Manley Law School and the Grand Palazzo Hotel were consistent with the standards normally used in the region for such buildings. The designers selected were experienced architects and engineers, and acceptable building codes were used for the design of the structures. However, the owners were not provided with statements indicating the strategy to be used for mitigating the effects of the known hazards, nor were the owners aware of the costs and benefits of the measures that would have avoided the damage incurred.

f) The failures of these buildings occurred not because of the use of inappropriate building codes and standards, but because the consultants did not specifically ensure that all structural and non-structural components of the buildings be constructed to resist the known hazards.

g) The manufacturer and suppliers of the roof system of the Norman Manley Law School were not required by the consultants to carry out tests of the roof assembly to determine the resistance of the assembly to uplift forces. If this had been done, the suppliers and the consultants would have seen that the holding down mechanisms were inadequate to resist hurricane winds.

h) In private sector projects, such as the Grand Palazzo Hotel, the emphasis is on effective management of the process of design and construction and on timely conclusion to the project implementation cycle. Record keeping is not the main concern of the owners, and the project is entrusted to competent professionals working in accordance with agreed deadlines and processes. Insurance coverage is considered to be a priority, but the design professionals are expected to carry out their work in accordance with the codes ("minimum standards") acceptable to the local regulatory body. This was the situation with the Grand Palazzo Hotel, but the regulatory body would not normally be concerned with the design of non-structural components. The duty of the regulatory body is to ensure compliance with the building regulations and with the building permit issued for the hotel.

The additional costs required to mitigate the damage suffered by the four projects studied varied from less than 1% of the original project cost to less than 12 % of the original cost as follows:

The costs of reconstruction as a percentage of the original costs were estimated to be:

These results show that if the mitigation measures were taken at the time of the original construction there would have been significant savings over the costs of reconstructing the facilities. If the mitigation expenditures required had been met, the damage incurred by these facilities would have been minimal, and would not have led to the loss of use of the facilities during the periods of reconstruction. In this paper the costs attendant upon the non-use of the facilities have not been measured. These costs are important in judging the benefits of the mitigation strategy for the project being considered.

While the CDB appraisal reports for the Dominica Deepwater Port and the St. Lucia bridges considered the stream of benefits from the use of the facilities, the appraisals did not specifically consider the strategy for hazard resistance or the costs of rebuilding the facilities in the event of major damage by hurricanes or floods. If this had been done, the estimated costs of the measures required to avoid the damage may have been found to be a necessary component of the capital costs of the project.

The costs of improving the resistance of the Normal Manley Law School and the Grand Palazzo Hotel to the hurricane forces that caused the damage to the buildings were minimal, and should have been incurred. It is evident that the owners were not advised about the need for the additional expenditures to prevent the losses sustained. The mitigation costs required to improve the resistance of buildings to hurricane forces is generally very small, and this study has confirmed this.

The following recommendations are considered to be necessary requirements for the development of a project designed to resist natural hazards:

3.1 Consultant contracts

Consultant contracts should require the consultants to carry out a pre-investment study, which includes, as appropriate. The project parameters shown in Item 2, Table 4.

This continuity of design and construction is desirable as the detailed design should be based on the decisions on hazard mitigation made in the preliminary design.

3.2 Consultant’s reports

a) Monthly reports by the consultants on the progress of the project are necessary to allow the owner and the financing agency to monitor the project. The reports should contain information on:

b) There should be a final report by the consultant on the construction of the works.

3.3 Building and infrastructure standards

a) The use of appropriate building codes and standards such as the OECS building codes and the Caribbean Uniform Building Code for buildings should be mandated by legislation.

b) Appropriate infrastructure standards for projects such as roads and drainage, bridges, ports, airports, and water supply and sewerage should be developed, so that construction of new infrastructure and maintenance and reconstruction can be based on firm requirements for hazard mitigation and durability.

3.4 Hazard mitigation

a) A hazard resistance strategy and the consequential costs and benefits should be an important project design requirement in the same way that environmental considerations are now integral parts of project documents.

b) The principles of hazard resistant design must be incorporated in the project development at the earliest possible stage so that all design and material choices made in the detailed design and subsequent construction can be based on the decisions taken and approved by the client and by the financial agency.

c) Reconstruction of lifeline facilities such as main road bridges must be carefully thought out even where the urgency demands hasty action. The ministry or institution involved in the reconstruction after damage by an extreme natural event must insist that the consultants or in-house engineers responsible for the design of the works develop long term plans (or strategy) to enable the facility to resist the known hazards. The contracts for the consultants engaged in such work should contain the provision that the design of the emergency work must be compatible with the need for hazard mitigation.

d) The owners should be made aware of the risks associated with the design concepts at the earliest possible stage in the implementation process, so that informed decisions can be taken. The consultant should explain the nature of the risks and the costs and benefits of the hazard mitigation strategy being recommended.

3.5 Review consultant

The owner/government involved with the planning and construction of an infrastructure project should appoint a review consultant. The review consultant shall report to the owner on the effectiveness of the hazard mitigation strategy being recommended and shall keep the project under review by examining the plans and periodically inspecting the construction.

3.6 Maintenance

Maintenance of important facilities including institutional buildings, waterways and bridge structures must be a mandatory requirement in the strategy for hazard mitigation.

3.7 Appraisal

The appraisal report of the financing agency should take into account the risks of damage to the physical assets being appraised by extreme natural events. The costs and benefits of the hazard mitigation strategy proposed should be factored into the stream of costs and benefits when calculating the rates of return on the project.

The Dominica Deepwater Port is in Woodbridge Bay just outside the city of Roseau. The Government of Dominica constructed the facility so as to handle its exports of bananas more efficiently and to lower the handling costs of imports. The final designs for the port, deemed to be affordable, represented between 50% and 60% of the initial scope of the works and consisted mainly of:

a) A 500' long marginal wharf for ocean-going vessels.
b) Berthing platform (wharf) for inter-island schooners.
c) Reclaimed area of about 5 acres (top elevation +9').
d) Revetment of boulders and reinforced concrete wall for protecting the reclaimed area.
e) Approach trestles between the main wharf and reclaimed area.
f) Transit shed of 10,000 ft^2.
g) Banana shed of 30,000 ft^2.
h) Pavements, fence and a fresh water system.

Construction of the port started in September 1974 with financing from the Caribbean Development Bank (CDB), using funds supplied by the United States Agency for International Development (USAID) and the Government of Dominica. By March 1976, the wharves, reclaimed area, revetment and approach trestles were completed. The construction of the buildings and other ancillary works was completed in 1978.

Consultants were appointed by ODA in 1969, to carry out the preliminary design of the port facility. CDB criteria was used for the selection of the consultants for detailed design and construction inspection. Their qualification and training were acceptable and the TOR were well defined. The consultants were committed by their terms of reference to pay due regard to the results of the soil investigations and hydrographic and land surveys, and to the danger of hurricanes, earthquakes, local swell and other weather conditions. The preliminary design and feasibility study were submitted by the consultants to the GOCD in 1971.

The consultants submitted 2 preliminary proposals for protecting the reclamation, one of which included a rubble bank with the front face in the tidal zone and above, protected with patented precast concrete blocks surmounted by a small wave wall (top elevation 11.5 feet), which would act as an edging kerb. This proposal was not accepted because of costs. The design agreed on was a based on constructing a rubble bank of stone of 20 150-lb. weights, on which would be placed gabions. A wave wall would be constructed.

The Delft Hydraulics Laboratory in the Netherlands did a study of the wave conditions. The Delft Laboratory was commissioned by ODA to carry out a study of the wave conditions of three islands in the West Indies including Dominica. In June 1972, a copy of their report was forwarded by the ODA to the GOCD. The GOCD in turn forwarded a copy of the report to the UK engineering consultants, just after the preliminary design and feasibility study was completed.

A copy of the Delft report could not be found, but references to it suggest that some of the conditions for Dominica identified in this report were :

After reviewing the Delft study, the consultants reported that they could find no reason to amend or change any of the conclusions and recommendations in their just completed engineering and economic (feasibility) study. A wave height of 6' had been used for the design of the marine structures. The consultants confirmed that in designing the new works—both the main structure and the fendering system—account must be taken of the probable maximum wave height given in the Delft study. The wind and wave conditions forecast in the Delft report would be taken into account in the design of the jetty.

The GOCD accepted the consultants’ preliminary design and costs and submitted this design to the CDB with their application for financial assistance.

Wind load pressures for the design of the transit shed were determined in accordance with the current (BAPE) wind code approved by CCEO, and SEAOC Zone 3 recommendations were used for the engineering designs to provide structural resistance to loads generated during earthquakes. The basis for the design of the banana shed, financed by CIDA, could not be ascertained.

The above wind code would have catered to a category 3 hurricane with wind speeds of about 111-130 mph (damage = extensive).

According to the HURDAT data base, compiled by the US National Hurricane Center, three category 3, four category 4 and one category 5 hurricanes passed through a 2-degree square centered on Dominica during the period from 1886 to 1992.

The final designs, which were based mainly on the preliminary designs used for the feasibility study, were carried out in 1973 and 1974 by a joint venture of a local engineering firm and an engineering firm from the USA. The joint venture had to submit pre-qualification information (in November 1972) as well as a technical & fee proposal (in March 1973), according to CDB criteria, before being selected to carry out the work.

The rules governing the use of the USAID funds, which the CDB used for the financing of the port, provided for the procurement of goods and services from eligible sources as listed by the USAID. The consultants who were appointed by ODA for the preliminary design work on the port were not eligible for selection under the USAID rules. The CDB did not request a waiver from USAID to allow these consultants to be used for the final design of the port.

The final design, in which the gabion wall was replaced by large boulders, was reviewed by the GOCD and the CDB for compliance with the accommodation requirements, but not for its engineering content.

The estimation and comparison of construction and other costs would have to take into account the variation of the EC $ to US$ currency rate, and the annual inflation rates between 1975 and 1982.

The currency rate during the construction of the port in the period from 1974 to 1976 fluctuated from 1.7857 to 2.65 EC $ to one US$. For comparison, all costs will be converted into 1975 (middle of the construction period) dollars based on the following conversion factors:

Year	Rate EC$:US$	Comments
1974	1.79	Start of the construction period
1975	2.22	Middle construction period—reclamation + marine works
1976	2.65	End construction period—reclamation + marine works
1977	2.66
1978	2.67	End construction period—buildings, ancillary works
1979	2.68
1980	2.69
1981	2.70
1982	2.70	Start of reconstruction period
1983	2.70	Middle of reconstruction period
1984	2.70	End of reconstruction period

a) For the initial construction work carried out on the reclamation and marine facilities, in the period from 1974 to 1976 (in EC $):

Physical facilities using tender rates	7,837,755
Escalation of labor and materials	762,878
Altering scope of works	30,770
Unfavorable rate of exchange	1,577.080
Liquidated damages	(46,000)
Professional fees (pre + post)	1,220,080
Sub-total	11,382,563
Converted to $US at an average rate of $2.22 EC to $1.00 US	5,127,300

b) For the buildings and other ancillary work carried out in the period from 1976 to 1978:

Transit shed - 10,000 sq.ft	420,300
Banana shed - 30,000 sq. ft	980,800
Ancillary works	358,600
Sub-total	1,759,700
Professional fees (pre +post) @ 4%	70,388
Converted to $US at a rate of EC $ 2.67 to US $ 1.00 and at an annual deflator of 12%	548,500

The total 1975 costs for the initial port facility, using the above computations, would therefore be US $5,675,800.

Expenditures were monitored monthly by CDB and GOCD. A full time project manager was appointed by GOCD and a full time resident engineer adn assistant were on site for monitoring of the construction.

The appraisal of this project was based on the savings that will accrue from the more efficient handling of exports (mainly bananas) and imports. The lighterage system used was expensive and caused damage to the fragile fruit on which the economy of Dominica depended. Shipment of bananas directly from the pier into the ship’s hold would reduce both handling time and damage potential. However, as such savings were relatively small and depended on the quantity of fruit exported, the capital and operating costs of the new port had to be reduced (if at all possible) to the levels that could be sustained by the estimated savings. This was the basic principle by which the GOCD had to design the accommodation requirements off the port.

No consideration was given to hazard mitigation and to the attendant costs and benefits of a strategy to reduce the vulnerability of the post to damage from hurricanes and earthquakes.

a) Approved by the GOCD, CDB and the consultant to GOCD.
b) Clearly defined in all of its essential physical features.
c) Costed at prices which the GOCD, CDB, and the consultant considered feasible.

On this basis, the capital cost of the port and the acceptance of the engineering design by the GOCD became the important determinants in the appraisal of the proposal.

Hurricane "David", a severe hurricane, passed over (or close to) the port in August 1979. The buildings and a portion of the revetment which protected the reclaimed area were severely damaged. The damage to the banana shed was more extensive than the damage to the transit shed. There was no evidence of any damage to the wharves. The approach trestles and the other ancillary facilities also experienced significant damage.

"David" was a strong category 4 hurricane (wind speeds of 131-155 mph, damage = extreme) . Published reports indicated that David had sustained winds with speeds that exceeded 160 mph and wind gusts of 200 mph. These wind speeds would more closely be associated with a category 5 hurricane (damage = catastrophic).

An assessment of the damage was carried out by the joint-venture consultants and, shortly thereafter, designs were completed for the repairs and reconstruction work necessary to make the port functional again. The restoration works included the use of four-ton concrete dolos to increase the resistance of the revetment protecting the reclaimed area against wave attacks.

The costs (in US dollars) as estimated by the CDB for the restoration/reconstruction work carried out in 1981 and 1982 are shown in Table A-1.

The total cost of US $3.933 million converted to 1975 dollars at 12% per annum is US $2.31 million.

The 1982 construction cost given above includes an extra amount (estimated at US $1.15 million out of the total estimated cost of shore protection of US $2.2 million) for the additional strengthening of the entire revetment using four-ton dolos.

Since the wharves were tested and found to be strong enough to resist "David", the additional costs for strengthening the rest of the port for a 15' wave (similar to the ones which developed during "David") would include:

a) Making the revetment more resistant to larger waves.
b) Raising the level of the reclamation from +9' to perhaps +15'.
c) Raising and strengthening of the approach trestles.

For the buildings to resist "David" force winds, they would have had to be designed for greater forces than the code indicated. This would most likely have led to the use of larger structural members, provided the existing building geometric configurations were to remain the same. The strength of the cladding would also have had to be increased and/or their supports and fixings placed at closer centers. The increased costs for the "David" design would, therefore, have been due to the increased cost of both the structural and non-structural elements.

While the original study carried out by Delft contained information needed to effect an appropriate design, the design consultants who were appointed in accordance with the CDB procedures should have carried out further studies to satisfy themselves and the GOCD that the design would be adequate to resist the wave forces generated by hurricane winds. It is estimated that the cost of the further studies would have been about US$ 30,000 (in 1975 dollars).

The construction cost given above for the reconstruction in 1982 includes an extra amount estimated at US $1.15 million (total estimated cost of shore protection, US $2,176,000) for the additional strengthening of the entire revetment using four-ton dolos, assuming that the reclamation level of +9’ would not be altered. Deflating this cost to 1975 dollars would give US$ 520,000, at the rate of 12% per annum.

a) Protective armor, raising the level of the platform: US $585,000
b) Strengthening of buildings: US $15,000
c) Further studies: US $30,000
d) Engineering fees and management: US $25,000
Total: US $655,000

The analysis of the costs of reconstruction, the original construction cost and the additional mitigation cost is given in Section B.

The problems that arose with the failure of the revetment and consequent failure of the ancillary works on the platform were due in large part to the selection wave heights used in the facility design. The designers were under pressure to maintain the lowest possible construction cost.

The appraisal of the project by the CDB showed that the project, as originally conceived, could not pay for itself from funds generated by the port and that both the financial rate of return and the economic rate of return were unacceptable at the time of appraisal. The length of the berthing platform was, therefore, reduced in order to reduce the capital cost of the project. In the appraisal of the port, no account was taken by the CDB of the problems of resistance of the structures to natural hazards and of the costs and benefits of hazard mitigation measures taken.

The Delft Hydraulics Laboratory, which had been engaged to study the heights of the waves which will affect the port site, submitted their report in June 1972. The consultants who were engaged in 1969 to carry out the preliminary design had recently completed their work and considered that their design concept with respect to the proposed heights of the reclamation need not be altered, despite the Delft study’s indication that significant wave heights of 15’ are to be expected once every ten years. The consultants confirmed that the design of the structure would take into account the effect of the high waves predicted.

The siting of the platform at an elevation of +9’ and the destruction of the rubble bank revetment were considered to be major causes of the failure of the reclamation. The berthing structures themselves were not damaged.

a) All studies to determine the hazards that would affect the design of the facility should be carried out under the responsibility of the design consultant.

b) The appraisal of infrastructure projects should take into account the need for resistance to extreme natural events in the same way that the maintenance of the facility and the impact of the proposed development on the environment are included as factors in the appraisal.

c) A review consultant should be appointed by the owner to review conceptual designs and studies to ensure that the designs show effective methods of resisting extreme natural events. The review consultant should also carry out periodic reviews of the construction.

d) For the Dominica Port, the cost of reconstruction was relatively high, about 41% of the cost of the original port. Most of this cost could have been avoided if the designs had taken into account the results of the Delft study and if the owner had engaged a review consultant to provide advice on the effectiveness of the design.

e) The additional cost of the facilities, if they had been designed to withstand the forces from hurricane "David", would have added 10 to 15% to the original cost. This additional cost would not have been accepted by the CDB without taking into consideration the benefits from the mitigation measures that would have led to the increased cost.

The Norman Manley Law School (NMLS) was established in 1973 by the Regional Council of Legal Education and provides a two-year Certificate in Legal Education for graduates with the LLB Degree. The building was constructed in 1974–1975 at a cost of about US $685,000. Financing was provided by the Government of Jamaica. The building houses the law library and lecture halls.

The project consultants are a well-established firm of architects and engineers and were chosen to carry out the detailed design on the basis of a design competition. The building is of reinforced concrete and concrete block masonry with a steel space frame roof on which is placed proprietary "tectum" deck planks and ¾" mastic asphalt waterproofing. The floor area is approximately 7,000 square feet (650.3 m²).

The contract for the manufacture and installation of the roof had been awarded to a competent firm of suppliers of proprietary roof systems. The consultant had provided the wind and other loads to be resisted by the roof structure to the manufacturer, which the manufacturer was instructed to use for the design of the space frame and the roof covering.

The consultants who won the design competition for the building carried out the design of the project. Preliminary studies, including the conceptual design, were carried out to determine the accommodation requirements of the Law School. No specific studies were carried out to determine the type and intensity of hazards which would affect the building.

The building was designed in accordance with the British codes of practice and in accordance with the building regulations of the Kingston and St. Andrew Corporation. The design for the roof was in accordance with the British standard code of practice for wind loading, CP3 Chapter V: Part 2:1972. The SEAOC recommendations for earthquake loading were used as the basis for the design of the building to resist earthquakes of the intensity projected to affect Jamaica. The codes used were satisfactory if implemented properly.

The University reviewed the conceptual design to determine whether the design met the accommodation and aesthetic requirements of the University. There was no specific mention of a hazard mitigation strategy in the consultants’ report, but the University staff interviewed indicated that it is normal for the engineers employed on building projects to design the buildings in accordance with the requirement of the building codes used in Jamaica. There is no record that the design was checked for its conformity to the building codes.

3. Final designs

The consultants who carried out the original conceptual designs were engaged for the preparation of the final detailed designs and for the supervision of construction in accordance with the contract of the Royal Institution British Architects.

The University reviewed the final designs before a construction contract was let. There is no evidence that any changes to the structural design to improve the building's resistance to natural hazards were made as a result of the review. The consultants carried out normal inspections and the GOJ reportedly inspected the construction.

Hurricane "Gilbert" (September 12, 1988) was the extreme natural event that caused failure of the buildings. It was reported that the hurricane produced winds in excess of 145 miles per hour.

The roof was badly damaged. The roof was constructed of steel space frames on which were placed "Tectum" deck planks and waterproofing of ¾" mastic asphalt. The roof planks were pinned to the roof, but according to the consultant not all of the planks were so fixed. Post-"Gilbert" evaluation indicated that the damage was due to inadequate fixing of the Tectum deck planks to the supporting steel roof members. The failure of a clerestory window contributed to the ingress of the wind and to the uplifting of the roof deck planks.

It must be noted that the structure of the roof did not fail and there was no other damage to the building. Fortunately, the librarian had secured some of the documents before the hurricane and, consequently, the damage to contents was minimal.

The ingress of the wind through the louvers added to the uplift forces of the wind and caused the damage to the roof of the Phillip Sherlock Centre for the Creative Arts (PSCCA) building.

5. Reconstruction

The University employed a project manager to oversee the reconstruction activities. As many buildings were damaged, the principal task of the project manager was to coordinate the reconstruction and to ensure speedy occupation of the damaged buildings. The terms of reference of employment of the project manager could not be found, but it seemed clear from discussions with University personnel that no firm instructions were given regarding the need to ensure hazard resistance in the reconstruction efforts.

The design and detailing work needed for the reconstruction were carried out by the original consultants, who also inspected the work in accordance with their contractual obligations.

In the reconstruction exercises for the buildings that suffered under "Gilbert", only partial structural design changes were made to the roof systems, reportedly because of financial constraints and limited time for re-occupation. In the NMLS building, the reconstruction work was mainly to restore the decking and waterproofing to the roof and the redecorate as necessary. The fixings to the deck planks were improved and the waterproofing re-laid. The consultants stated that, in their opinion, the roof of the NMLS building was now resistant to wind forces of "Gilbert" strength.

The cost of the reconstruction for the NMLS building was given as US $90,000, but the University took the opportunity to carry out some deferred maintenance, so the cost of repair due to the hurricane damage was overstated.

The consultants stated (as reported in 2.2) that they had used the British wind code and the SEAOC earthquake recommendations for the structural design of the building. The structure of the building was not damaged by the hurricane forces. However, the fixing of the roof deck planks—a critical item for lightweight roofs—was not adequate to resist the uplift forces generated by "Gilbert". The consultants had supplied the manufacturer of the proprietary roof with the appropriate wind speeds and uplift forces, but it would appear that the installation details were not properly checked.

During the reconstruction, the consultants took care to ensure that the fixings of the roof deck planks had been improved and that each plank was securely fixed to the supporting steel frame. For the other buildings damaged by "Gilbert", there was no record that there had been a change in the fixing details to the light weight roofs, although the roofs have been repaired.

The consultants for the NMLS building stated that they had the information required for proper design of the building. The wind code used is considered to be adequate for buildings in Jamaica and the earthquake code used is the standard code used by all Jamaican structural engineers. The only extra studies and engineering that would have been required would probably have been the testing of the roof assembly for resistance to hurricane wind forces, and the extra engineering in developing and checking the fixing details for the roof deck planks.

It is estimated that a total of US $13,000 would have allowed the consultants to carry out any testing as necessary and to pay for the installation of the extra fixings. This estimate of increased cost includes US $5,000 for studies and engineering supervision, and US $8,000 for the supply and fixing of extra fastening mechanisms for the roof deck planks.

The NMLS building suffered damage because the roof deck planks were not all securely fixed. Very often the responsibility for the details of non-structural elements is not made clear in consultant contracts. It is normal for the structural engineer to be responsible for the roof structure and to supply the manufacturer of the roof frames and covering with the necessary information about wind speeds and uplift forces. In this case, it appears that the consultants were not aware that the roof deck planks had not been adequately fixed to resist the uplift forces generated by "Gilbert".

The University has recently improved its management of new construction on the campus. This will go a long way towards ensuring that construction records are available at the University. Unfortunately, the records of the NMLS were not readily available and no as-built drawings were found at the University. The staff now concerned with the maintenance of the facilities should have all drawings and documentation of the buildings to be maintained.

While a variety of useful information was collected from the limited project documents that were located, the information was not readily accessible. Little or no information was available at the University on specific aspects of the projects such as natural disaster preparedness/designs for disaster mitigation and how these would be treated or implemented in the project. There is no central office/location to provide access to documents.

The consultants who were engaged in the design of the Normal Manley Law School were very cooperative and supplied the technical information on the construction of the building and especially of the roof. This information should have been available at the University and it is recommended that the University set up the necessary mechanism to allow for the storage and retrieval of such vital information.

a) There must be a greater involvement of the owner in the design process to ensure that the hazard mitigation strategy used by the consultants is adequate to resist the known hazards.

b) For efficient implementation of the project, there is need for roles of the organizations involved in the project to be defined, and the specific responsibilities of the executing agency, the financing agency and the owner for hazard mitigation should be clearly documented.

d) A pre-investment study should be submitted by the consultants to the owner as a formal document. The study should include the strategy for hazard mitigation, the building codes to be used, the principal materials of construction recommended for the building, and costs and benefits of the measures taken to resist the known hazards.

e) The contract for structural engineers engaged on the project must include the requirement that the engineers would be responsible for checking the resistance of non-structural elements to hurricane and earthquake forces. This would require a specific change in the normal terms of engagement of structural engineers, but the failure of non-structural elements such as windows, doors and roof covering often leads to major failure of the building and consequential damage to its contents.

f) The reconstruction of damaged buildings must be taken seriously and every effort made to correct errors in the design details, which led to the failure.

g) Project managers employed for the reconstruction of buildings damaged by extreme natural events should be given firm instructions to ensure that the design and construction systems being used conform to the applicable building codes and that a hazard mitigation strategy is in place.

h) As built drawings must be submitted by the consultant to the owners at the end of construction of the original facility. Drawings of any changes to the building made during the reconstruction or after major repair must also be submitted to the owner.

i) For all important buildings, the owner must engage a review consultant to review the design drawings and to advise the owners on the effectiveness of the hazard mitigation strategy employed by the design consultants.

1. Original project description

The two St. Lucia bridges studied are the Troumassee bridge on the Castries to Vieux Fort main road and the Caico bridge on a secondary road to Millet. Both of these bridges are important to the economy of St. Lucia.

1.2 Troumassee bridge

The Troumassee bridge spans the Troumassee river and is on the connector road from the international airport at Vieux Fort in the south of St. Lucia. The Caico bridge spans a smaller river allowing transport of bananas from the Millet area to the sea port at Castries.

In 1972, the Caribbean Development Bank (CDB) was requested by the Government of St. Lucia to assist in financing the reconstruction and/or construction of six main road bridges in St. Lucia. The Troumassee bridge was one of the six bridges to be reconstructed. The scope of work was well defined and CDB criteria was used in the selection of consultants. Consultants were appointed to carry out a pre-investment study of the six bridges and submit a report to the Government of St. Lucia and the CDB. The study was completed in February 1973 and the CDB agreed to assist with the financing of the project on the basis of the study.

The Troumassee River widens into a flood plain where it crosses the main road. The dry weather stream sizes given in the description below bear no relation to the size of the river under flood conditions. The basic description of the Troumassee bridge is:

Watershed area: 7,620 acres
River width: 90 ft
River depth: 7 ft
Flood level: 335.46*
Bridge deck level: 339.46*

*Note that levels refer to a benchmark on the main road and are not necessarily representative of heights above mean sea level.

The bridge as constructed consists of two spans of 68’ each with reinforced concrete abutments and center pier. Gabion protection was supplied to the banks of the river to prevent the fill behind the abutments from being washed way.

The consultants in their completion report also recommended strongly that annual maintenance of the bridge structure and of the gabions be carried out. The chief engineer of the Ministry of Communications and Works decried that maintenance funding was low and that adequate maintenance to the bridges could not be carried out.

When the survey was carried out in 1973 for the rehabilitation of the Troumassee bridge, there were no reports of the flood overtopping the bridge deck. The consultants were thereby assured that the hydraulics of the existing waterway were adequate, although there was some scouring occurring at the upstream side of the north abutment. There have been reports, however, that the floodwaters have reached the deck of the bridge.

1.3 Caico bridge

The Caico bridge that was damaged by floodwaters was constructed in 1995. It replaced a bridge that was damaged by the 1994 flood. The superstructure was a Bailey bridge supplied by the British Development Division (BDD) and installed by the Ministry of Communications and Works of St. Lucia. The bridge was a single span of about 50 feet and the superstructure rests on existing stone abutments. This bridge was examined in some detail, as the original bridge constructed more than 20 years ago was destroyed by the 1994 floods. The area served by this bridge is an important banana producing area. The chief engineer reported that adequate wing walls and riverbank protection were not constructed for the reconstructed bridge due to lack of funding.

There is no hydrological information or hydraulic design data for the bridge. The bridge was constructed hurriedly to provide access to Castries for banana traffic and for normal vehicular traffic from the Millet area.

The failure of the bridge is a result of the lack of wing walls, footing protection, and stabilization of the riverbanks. There was no redesign of the bridge apertures. The increased flow resulting from development upstream of the bridge and the heavy rainfall meant that the original span of 50 feet was inadequate to accommodate the floodwaters. The result was the destruction of the approaches, undermining of the bridge abutments, and overturning of the bridge superstructure during the tropical depressions of 1996.

The currently planned reconstruction work includes the replacement of the bridge with a Bailey bridge with a span of 80 feet and adequate wing walls and river training works. The cost of this work is estimated (1997) to be US $317,000.

2.1 Troumassee bridge

As stated in 1.4, the consultants for the Troumassee bridge prepared and submitted to the owner, the Government of St. Lucia (GOSTL) and to the CDB, a pre-investment study which contained relevant information on soil investigations, hydrology and hydraulics at the bridge sites, design, costs, and alternative systems.

The design flood was calculated using the best information on rainfall intensity and frequency available at that time from the Mamiku station. In calculating the design flood, the consultants assumed that the total area of St. Lucia is small enough so that the Mamiku figures would be representative.

Statistical procedures were used to produce rainfall intensity duration curves for 2, 5, 10, 15 and 25-years, based on the maximum daily rainfall figures at Mamiku. From these curves, the design intensities for a 25-year return period were arrived at assuming the Mamiku area figures applied to all sites studied. The design intensities were different for each site since they are affected by the time of concentration, which varied from site to site. For a 25-year recurrence period, a one-hour rainfall intensity of 2.88" was used. This was equivalent to a maximum 24-hour intensity of 11.45".

The pre-investment study provided detailed calculations for the stream flows and scour depths for the bridges studied, and the conceptual design was based on the conclusions of the study. The GOSTL and the CDB accepted the pre-investment study report and conceptual drawings. The consultants were engaged to carry out the detailed design of the bridges based on the results and recommendations of the pre-investment study.

2.2 Caico bridge

3.1 Troumassee bridge

The final designs for the Troumassee bridge followed the recommendations of the pre-investment study. The construction contractors, however, recommended alternative construction details, which were accepted by the consultants and by the GOSTL. The alternatives were:

a) Use of steel piles filled with concrete instead of greenheart timber piles.
b) Raising the pile caps above the river level to ensure construction "in the dry".
c) Providing gabion protection against loss of fill associated with raising of the pile caps.

In the consultant’s opinion, these amendments did not affect the ability of the structure to resist the forces induced by floods.

The designs of the bridge structures were based on B.S.153–"Girder Bridges" Part 3A–Loads. The bridges were designed to accommodate trucks of up to 22 tons in weight while this loading also makes allowance for the possibility of loading from more than one vehicle. Allowance was also made for the forces due to traction or braking of vehicles, as these forces usually have a significant effect on abutments.

The expected traffic density for which the bridge was designed was 400 to 600 vehicles per day. This criterion was also used for the design of the roads adjacent to the bridges.

3.2 Caico bridge

An engineer from the Ministry of Works designed the Caico bridge. The Caico bridge superstructure was installed on existing stone abutments. There was not a formal preliminary design, but the Bailey bridge was capable of accommodating the loads similar to those of the Troumassee bridge.

4.1 Troumassee bridge

The Troumassee bridge was constructed in 1975–1976 at a cost of about EC $375,000, including the cost of engineering and management. This cost was equal to about US $170,000 in 1975 at the 1975 exchange rate of 2.20 EC dollars to one US dollar. The consultants noted that the final cost of the construction was just 0.6% greater than the estimated costs as given in the pre-investment report.

The CDB monitored construction mainly for cost control and a resident engineer was on site for quality control.

4.2 Caico bridge

The cost of the bridge superstructure of the Caico bridge provided by the BDD in 1995 was not available. However, the estimate of the original cost of construction has been calculated based on cost parameters developed by the consultants of the Troumassee and other main road bridges in 1976 and inflated to 1995 costs. This cost of EC $150 per square foot of bridge deck inflated to 1995 cost gives an estimate of cost of the bridge in 1995 as EC $315,000.

Tropical storm "Debbie" struck St. Lucia during the early hours of September 10, 1994. Rainfall was reported as high as 15 inches over a 10-hour period and contrasted sharply with a monthly average of 12 inches for the month of September over the last five years. Heavy rains in the previous week had resulted in extensive soil saturation.

The rainfall is generally believed to be of an intensity that occurs once in every 50 years, although the consultants on the Roseau dam project have suggested that the storm was at least a one in 500-year event [5]. The rainfall was of sufficient intensity to cause financial loss estimated at EC $209.1 million, or 18.9% of the 1993 GDP.

The tropical depression of October 1996 produced about 8 inches of rainfall in 24 hours and caused damages to the bridges of about EC $1.8 million or US $685,000. The northern approach to the Troumassee bridge was washed away. The damage to the Caico bridge included the overturning of the one abutment due to the lack of adequate wing walls and the consequential damage to the Bailey bridge.

The Ministry of Communications and Works has estimated the cost of reinstatement of the bridges (including engineering and management) to be EC $320,000 (or US $120,000) for the Troumassee bridge and EC $860,000 (or US $317,000) for the Caico bridge.

The reconstruction necessary for the Troumassee bridge is the rebuilding of the river bank protection with gabions. The estimated cost of this work is US $120,000. The main problem with gabion protection is the need for maintenance. This was pointed out by the consultants in their completion report in 1976. Unless the gabions are maintained there will be no protection during the flood, when it is most needed.

The estimate of cost of the Caico bridge is based on the construction of new bridge with a widened span of 80 ft, and constructing adequate abutments and wing walls. A single span bridge is recommended as the stream under flood conditions carries boulders and branches of trees and other debris which can significantly obstruct the flow. This estimate of US $317,000 would compare reasonably well with the costs of the bridges constructed in 1975 when inflated to 1998 costs.

7.1 Troumassee bridge

It is estimated that an amount of no more than US $20,000 would have been needed to prevent the damage to the Troumassee bridge. The consultants in their recommendations to the Government of St. Lucia did recognize the need for protection of the riverbanks. As development on the watershed increases, the flow characteristics of the river may change. There should, therefore, be half yearly or yearly examinations of the expected flow conditions and examination of the need for protection of the banks. The prevention of silting apparently led to the decision to reduce the aperture of the bridge, which in turn caused the destruction of the north approach.

At the time of the construction of the Troumassee bridge, an expenditure of a further US $8,000 would have provided the Government of St. Lucia with more information on the effects of rainfall with a fifty year occurrence. This in turn may have led to further construction of the gabion protection. It must be recognized, however, that unless the gabions or any other form of protection are adequately maintained damage for tropical storms will ensue.

7.2 Caico bridge

This bridge constructed in 1995 after the destruction of the previous structure by tropical storm "Debbie" in 1994 was damaged by the 1996 tropical depression. The main problem as reported by the chief engineer of the Ministry of Communications and Works was the lack of adequate wing walls to channel the flow and the constriction of the flow caused by the 50 foot span of the bridge.

The recorded rainfall which led to the damage of the Caico bridge was about 8 inches on October 26, 1996. In 1973, the consultants working on the bridges financed by the CDB and GOSTL developed a graph of annual maximum one-day rainfall and rainfall frequency curves. These curves showed that the 24 hour rainfall of 12" is probable once in every 25 years. The design of the Caico bridge, if done in accordance with the information already at hand, would have led to a bridge which would not have been severely damaged by the 1996 tropical depression.

It is recognized that the bridge was constructed in haste to accommodate the traffic, and that at the time of construction it would not have been possible to increase the span and construct a single span bridge under these time pressures. However, if the wing walls had been constructed to ensure channeling of the flow and bank protection installed, the damage to the bridge would have been considerably reduced.

a) Research of existing information: US $2,000
b) Preparation of preliminary design: US $6,000
c) Preparation of design of wing walls and protection of the footings against scouring: US $7,000
d) Design of protection to banks: US $2,000
e) Construction of wing walls and footing protection: US $25,000

The general conclusions that can be reached have to do with the process of design and construction of infrastructure works rather than with the need for further information. The design of the waterways of the six main road bridges constructed in 1975/1976 was based, as given in the pre-investment report, on a 1 in 25-year storm. The examination of the condition of these bridges showed no significant damage from the floods of 1994 and 1996.

The information on stream flows under storm conditions was available to the Government of St. Lucia in the pre-investment report in 1975, and although this information should probably be updated, the designs based on the information of 1973 proved to be adequate to withstand the effects of the floods and tropical depressions.

The process of design must be improved so that the engineers employed on the tasks of constructing infrastructure facilities be encouraged, or required, to examine all pertinent information and to submit for discussion and approval a conceptual design which would address the hazards. Even where construction has to be immediate, the engineer should still be required to look at the long-term needs of the facility and to deign the remedial or reconstruction works accordingly.

Maintenance of waterways has been shown to be an important aspect of hazard mitigation. Adequate funding for this activity is always difficult to secure, but non-maintenance often leads to dramatic and expensive failures.

The retrospective look at the design and construction of the Troumassee bridge was made possible by the existence of a pre-investment study and a completion report of the project. These documents are invaluable to the process. It is strongly recommended that financing agencies and governments constructing infrastructure projects require their consultants to prepare pre-investment studies which show all the steps taken to mitigate the known hazards, and completion reports of the construction.

The Grand Palazzo Hotel was constructed by Pemberton Resorts in 1992. The hotel is situated at Great Bay in St. Thomas.

a) Many of the windows of the reception building were fitted with louvered shutters, that form protection from hurricane winds. It was not evident that hurricane shutters were planned for any of the other buildings.

b) Some of the roofs are of concrete construction and these are secure from high winds. Most of the roofs have clay barrel tiles as their covering. It is understood that each roof tile was fixed with screws to the supporting timbers structure. This frequency of fixing along with the slope of the roof would render the tiles safe from removal by high winds.

c) The timber rafters, ridges, hips and purlins, which form the main supporting elements for most of the roofs, are connected to each other and fixed to the concrete structures with hidden connectors. The connection details would provide adequate resistance to high winds.

d) The examination of the drawings showed that the fixing details between the window and doorframes and the supporting building structures were sound. However, the structure of the large external doors of the main dining room does not appear to be strong enough for the wind criterion of 130 mph which was used for the design of the structure.

There is no information on the studies carried out to determine the frequency and intensity of the known hazards. It is normal for structural engineers, in designing the structures of buildings, to rely on the studies carried out by the developers of the building codes and standards to be used for the specific area in which the building is to be constructed.

The principal document controlling all building in the United States Virgin Islands (USVI) is the Virgin Islands Zoning, Building and Housing Laws and Regulations. This document was published in 1984 with a supplement dated 1987.

Following hurricane "Hugo" in 1989, the USVI decided that all Government buildings must comply with the Uniform Building Code (UBC) published by the International Conference of Building Officials of the United States. The UBC 1988 edition formed the basis for design of the Grand Palazzo Hotel. The following criteria were used:

a) For earthquake loading: Zone 4. This is equivalent to what is known as the "super zone" in California (the worst areas in the State).
b) For wind loading: basic wind speed of 130 mph. (fastest mile). This is considered to be very severe for conventional commercial facilities.
c) For flooding: a one-in-a-hundred-year intensity of 9.5 inches per hour.
d) For fire: a one-hour rating.

After hurricane "Marilyn" all major buildings were required to comply with UBC. This legislation was enacted late in 1995.

The above listed design criteria would appear to be satisfactory and in some cases more than satisfactory for the design of the structure to resist hurricane hazards.

Conceptual designs showing the accommodation and aesthetic requirements are normally provided to the owners of hotel developments for their review, but there is no evidence that hazard resistance requirements were made a part of the conceptual design.

Although the design of the project was carried out by a comprehensive team of established consultants, no information was available on the criteria used to select the consultants.However, no important aspect of the works was left to chance. Specialist consultants were employed to design most aspects of the project with terms of reference which would be regarded as normal good practice for a major project.

The monitoring of the project and the management of the construction process was carried out by an integrated team of professionals including the structural engineer, the architect and the mechanical and electrical engineer. There is no evidence, however, that the financing agency was involved in the monitoring of construction. In summary:

a) The design consultants carried out intermittent inspections with staff who were not in full-time attendance on the site. Such inspections would have provided reasonable assurance that the works were being carried out in accordance with the drawings and specifications.
b) The USVI Department of Planning and Natural Resources provided inspectors to check the critical stages of construction in accordance with the building regulations.
c) The construction managers, as part of their certification of payments procedures, would have been concerned with quality as well as quantity.

There is no information as to whether the final designs departed from the conceptual or preliminary designs. It is normal for changes to be made to the conceptual designs of facilities such as a hotel, but the basic structural scheme should not have been changed significantly.

Design Stage I services include the required investigation into the project and lead to the consultant’s formal proposals as documented in drawings and reports. Judging from performance of the facility and other indicators, the incorporation of code requirements was imperfect in Design Stage I.

Our information is that the construction of the hotel was originally financed by private banks with Barclays Bank as the lead institution. In some areas, notably in Trinidad and Tobago, Barclays Bank Group may monitor the construction activities of major projects financed by the bank and its subsidiaries. In one such project, the terms of reference for the employment of engineers included:

This list of requirements for the examination of pre-construction activities would ensure that the project will be designed in accordance with principles of hazard resistance as recommended by the codes and standards being used. While this review may not have been carried out in the Grand Palazzo Hotel project, Barclays Bank is obviously conscious of the need for such reviews to provide the bank with comfort that the facility has been appropriately designed and that there should be no major implementation problems.

Our information is that the total cost of the facility including the cost of furniture and equipment was US $28 million. Of this, about US $4.0 million would have been the cost of professional services for design and management.

The event that caused damage to the facility was hurricane Marilyn on September 15–16, 1995. This was a category 2 hurricane with winds gusting at 112 knots.

The total losses of the USVI were estimate at US $1,800 million or 1.5 times the GDP of the USVI. The total insured losses were US $650 million.

a) Loss of roof structure of the recreation building.
b) Loss of the sail cloth roof of the dining area.
c) Roof tiles broken probably by impact from flying debris.
d) Some collateral damage reported to windows, contents, grounds, and swimming pool.
e) Failure of some roof anchorage details (especially for the balcony roofs).

The reconstruction work needed was the replacement of the damaged roof, rebuilding the reception building’s roof, replacement of damaged furniture and equipment, and work on the grounds and outside areas.

There would have been no need for further studies to determine the intensity and frequency of the hazard. The building code and standard used was adequate. There was need, however, for further engineering work on the design of the roof and probably on the columns, which appear to be undersized. It is estimated that a sum of US $2,000 would have been required for the necessary detailing work. However, the fee paid to a structural engineer would normally allow for the design to be carried out in accordance with the required codes and there should, therefore, be no additional costs of design. The amount of US $2,000 is the estimated cost to the engineer for developing the design details.

The additional cost of construction required to mitigate the damage incurred is estimated at US $25,000 mainly for the structure of the roof of the reception building. This sum does not include the cost of improving the strength and resistance of the sail cloth roofs in the dining area. Sail cloth roofs can be made to resist high winds, with more frequent fixings, heavier material and other design changes. The cost of improving the structure where required to comply with the earthquake resistance standards has not been computed.

* Note: Not including the cost of possible replacement of the fabric roof of the dining area with a more resistant construction.

The implementation of the project was carried out with professional groups employed for each element of design, procurement, and construction and installation. The organization charts show that the owners were deeply involved in the architectural design, landscape design and procurement aspects of the project. There is no evidence that the owner was involved in the structural engineering design or in the strategy for hazard mitigation.

The major weakness in the organization would appear to be the lack of a review consultant for examining the details of the architectural and structural design to ensure compliance with the required codes and standards.

This report has concentrated on the resistance of the building to hurricane forces as hurricane "Marilyn" caused some damage to the hotel. However, the building must also conform to the USVI building regulations with regards to other hazards such as fire and floods.

The UBC provides standards for fire resistance to which the building must meet. The concrete construction of most of the blocks will meet the UBC standard except that in the "back-of-the-house" area the structural steel should be covered with appropriate material to provide a 2-hour resistance grading [6]. A one-hour fire resistance is considered to be marginal. The facility is also supplied with sprinkler systems.

It is understood that a specialist consultant was employed to review the sprinkler system.

The site of the building has a natural drainage pattern and the built drainage systems are designed for a very high intensity rainfall. There should, therefore, be little damage from heavy rainfall.

a) The efficient implementation of a project depends on the creation of an organization in which the responsibilities of each member are clearly identified and in which each member is professionally qualified and experienced in the tasks to be carried out. Such an organization was developed for the Grand Palazzo Hotel project.

b) The owner should be involved in the decisions affecting the resistance of the hotel to natural hazards. The owner will then have the relevant information so that decisions on the choice of building systems and choice of materials to be used for the structure and for non-structural purposes can be made in full knowledge of the probable consequences.

c) Large glass windows and doors are vulnerable unless specially designed. The damage to glass windows and doors by flying debris can be prevented by the use of impact resistance glass, which is now being incorporated into codes and standards.

d) A review consultant should be engaged by the owner for the review of the structural and architectural designs and to report to the owner on the compliance of the designs with the codes and mitigation standards.

Items examined	Projects
	Dominica Deepwater Port	University Buildings	St. Lucia Bridges [2]	Grand Palazzo Hotel
Development of the project concept — the project document, proposed scope and benefits.	Information found. Benefits defined. Terms of reference required that consideration be given to hurricanes, earthquakes, and squalls. Appraisal document for reconstruction examined. CDB estimates the financial rate of return is 10%.	Project concept and scope defined as one of the requirements for a design competition. The relevant documents were not found at the UWI, but discussions with UWI staff and with the design consultants indicated that project was well defined.	Information on benefits not found. Pre-investment study report submitted for six bridges, including the Troumassee bridge being financed by CDB and GOSTL. Project scope well defined for CDB-financed bridges. Appraisal document not found.	Project scope reasonably well defined. Information on benefits found.
Arrangements for financing.	CDB/USAID, CIDA and GOCD. IBRD assisted with payment for studies after "David".	Government of Jamaica	CDB and GOSTL for the Troumassee bridge. BDD and GOSTL for the Caico bridge.	Private banks with Barclays Bank as the lead financing institution.
Choice of consultants - training and qualifications, briefing, terms of reference, remuneration.	ODA selected consultants for conceptual design and feasibility study. CDB criteria used for seletion of consultants for detailed design and construction inspection. Qualifications and training acceptable. TOR well defined.	Chosen on the basis of a design competition. The consultants chosen were well qualified and experienced in design and construction of buildings in Jamaica.	CDB criteria used for the Troumassee bridge. Qualifications and training acceptable. TOR well defined. Ministry of Works engineers designed the CAico bridge.	No information available on criteria used for choice of consultants. The consultants used were well qualified and experienced in the design of buildings. The consultant's terms of reference and responsibilities on the project were normal for this type of project.
Project monitoring by: Financing agency Government or owner	Expenditures monitored monthly by CDB and GOCD. Project manager appointed by GOCD. Full time resident engineer and assistant on site for monitoring construction.	Consultants carried out normal inspections and the GOJ reportedly inspected the construction.	For Troumassee bridge: CDB monitored construction mainly for cost control. Resident engineer on site for quality control. For Caico bridge: GOSTL engineers monitored construction. Bridge deck and superstructure supplied by BDD.	Consultants on site. Monitoring procedure well defined. No evidence that financing agency involved in monitoring construction although Barclays Bank in Trinidad and Tobago monitors construction of projects it finances. (see Appendix D) The consultants dealt directly with the regulatory body in St. Thomas. No evidence that the owners were involved in checking structural drawings for hazard mitigation.
Preparatory studies carried out by consultants to determine the nature and frequency of extreme natural events.	Preparatory studies were carried out by consultants employed by the donor, ODA. Studies of wave regime and frequency of squalls and hurricanes were carried out.	None carried out. Structural design based on the requirements of the relevant British codes of practice which were in use in Jamaica. Wind code used was satisfactory for use in Jamaica.	For Troumassee bridge: studies carried out by consultants. Pre-investment report gives hazard mitigation strategy. For Caico bridge: No preliminary study carried out. This bridge, constructed in 1995, was in immediate response to flood damage in 1994 of existing bridge.	No information on formal studies of known hazards. Physical planning and accommodation studies carried out. Structural design criteria based on the UBC code requirements.
Conceptual design and design criteria (both basic and derived criteria) used by design consultants.	Preliminary design and feasibility study submitted by original consultants. Design for facility as constructed developed on basis of discussions between CDB, GOCD and consultants. Design criteria based primarily on accommodation requirements and on financial and economic viability. Study on wind and wave regimes carried out, but results were not used in preliminary design.	Preliminary design approved by the university. Stated criteria not found. Although there is no evidence that hazard mitigation was included in the brief to the consultants, there was specific reference to resistance to hurricanes and earthquakes.	For the Troumassee bridge, design criteria stated. Conceptual designs provided in the pre-investment study report. No formal design was carried out for the Caico bridge.	No firm information. It is normal for the architectural consultants to prepare conceptual designs for hotel projects. Such designs would stress the accommodation and planning aspects of the project but not necessarily the resistance to natural hazards.
Was a formal design Stage I report submitted for approval?	No formal design Stage I was submitted by design consultants.	Yes, accommodation requirements discussed and approved by the university.	Designs provided in the consultant's pre-investment report.	No information. Stage I design would normally have been included in the conceptual design.
What codes were used in the design of the project?	There is no specific information on codes used for the design of the port structure. The wind code developed for the Barbados Association of Professional Engineers was used for the buildings. The consultants stated that US standards were used for the design of the berthing platforms and port structures and for resistance of the structures to wind and wave forces. The design of the berthing structures proved to be adequate to resist the high waves and winds of hurricane "David".	UK codes of practice and the KSAC building regulations were used for the planning and structural engineering. The codes used for the design of the roof structure were not stated, but the loadings were supplied to the manufacturer. It was a normal practice at that time for engineers in Jamaica to use the appropriate UK wind code for the design of buildings.	For the Troumassee bridge: There is no specific information on codes or standards used for the bridge design. The consultants used available technical information on the design of bridge structures to accommodate projected bridge loadings and stream flows. For the Caico bridge: No information on the design standards used was available.	Consultants indicated that applicable earthquake and wind codes were used in the design. The results however indicated that the actual work was not always in accordance with codes. (See appendix D)
Did the work carried out in design Stage II follow the decisions made in design Stage I.	Stage I decisions not specifically known. Hence no check was possible to determine whether Stage II decisions followed Stage I decisions.	Structural engineering design followed the Stage I design.	For Troumassee bridge: yes For Caico bridge: no formal stage I or stage II designs carried out.	Problems have been noticed in the Stage II design. See accompanying report - Appendix D.
Involvement of owner and donor agency in design decisions where resistance to natural hazards are concerned	CDB requested information from the original consultants on basic design decisions for loadings including wind and earthquake loads. No specific information was provided and there was no follow up by CDB. Owner involved in monitoring accommodation requirements. CDB very involved in design decisions for the reconstruction. Owner involved in decisions on hazard mitigation.	Owner very involved in design decisions affecting the planning and accommodation requirements. There is no evidence that the owner was involved in structural decisions affecting hazard mitigation.	For Troumassee bridge: CDB monitored design decisions in the pre-investment report but made no comments. No evidence that owner was involved in the technical decisions affecting resistance to natural hazards, except through discussions with the consultants and approval of the report and conceptual designs. For Caico bridge: No information on design criteria used. Owner involved in construction, but no specific study of hazard mitigation was carried out.	No evidence that financing agency was involved. Owner was involved in decisions affecting accommodations and facility requirements. No evidence that owner was involved in decisions affecting hazard mitigation.

Items	Dominica Deepwater Port	Norman Manley Law School – Jamaica	Troumassee Bridge - St. Lucia	Grand Palazzo Hotel - St. Thomas
Original project cost (year)	5,676,000 (1975)	685,000 (1975)	185,000 (1975)	28,000,000 (1992)
Reconstruction cost (year)	3,933,000 (1982)	90,000 (1990)	120,000 (1998)	5,600,000 (1995/1996)
Construction price deflator (per year)	12.0 %	7.9 %	5.9 %	1.8 %
Deflated reconstruction cost (year)	2,310,000 (1975)	28,800 (1975)	32,100 (1975)	5,308,000 (1992)
Reconstruction cost as a percentage of original cost	40.7 %	4.2 %	17.4 %	19.0 %
Additional mitigation cost (year): Studies Engineering Construction	30,000 (1975) 25,000 600,000	3,000 (1975 2,000 8,000	5,000 (1975) 3,000 12,000	0 (1992) 2,000 25,000*
Elements damaged	Port buildings Reclamation Access bridges Ancillary infrastructure	Roof covering Some furniture	The north approach of the bridge was washed away and the river banks eroded.	Roof structure and roof covering on one building Sail cloth roof over dining terrace Broken tiles, and other collateral damage
Reconstruction cost allocation Construction Engineering and management	93% 7%	78% 22%	86% 14%	85% 15%
Additional mitigation cost as percentage of original construction cost	11.5 %	1.9 %	10.8 %	0.1 %
Additional mitigation cost as percentage of reconstruction cost	28.0 %	45.0 %	16.7 %	0.5 %

Item	Phase	Work to be done	Responsibility
1	Project identification	Identify need and project parameters. Inputs and expected outputs to be assessed.	Owner with assistance of the financing agency
2	Pre-investment study	Studies to be carried out as appropriate: Demographic assessment and land use Identification of hazards and mitigation strategy including siting of facilities and shape of buildings to reduce the effects of the identified hazards Topographic surveys, hydrology, and oceanography Accommodation requirements Environmental impact assessment and strategy to prevent adverse environmental effects Design criteria to be established by consultants Preliminary choice of construction materials Alternative systems of design and construction Conceptual drawings to be prepared showing principal systems to be used; systems must be based on hazard-resistant principles Preliminary costings Outline program for implementation of the project Preliminary identification of benefits and determination of economic rate of return and financial rate of return	Owner/ review consultant with specific knowledge of similar projects to be employed. TOR for the engagement of the review consultant to be developed by owner with assistance of the financing agency TOR for pre-investment study to be developed with the assistance and approval of financing agency
3	Submission of study	Consultants submit study to the owner and discusses the design criteria, issues of hazard mitigation, environmental control and accommodation requirements.	Consultant.
4	Review of study	Special review consultant reviews report, plans and documentation and in particular ensures that the study recognizes the known hazards and that the conceptual design: Takes into account the strategy for mitigating the effects of the hazards Provides the accommodation required The review consultant advises the owner on compliance of the plan with the recognized principles for resisting the effects of the extreme natural events. Study is further reviewed by the owner and the financial agency.	Owner and review consultant
5	Proposal to financing agency	Owner submits formal request to financing agency for financial assistance, based on results of the pre-investment study and on the advice of the review consultant and including any amendments suggested by the financing agency.	Owner Financing agency
6	Project appraisal	Financing agency examines information available and requests further information if required. Appraisal examines the ability of the project to be self financing and to be of economic benefit to the owner. Project design to show minimum adverse environmental impact and maximum hazard resistance to known hazards, and must be in accord with acceptable design principles and codes.	Financing agency
7	Project approval	Financing agency approves project as appraised with standard conditions and special conditions on: Technical standards and codes to be used for the detailed design of the facilities Use of consultants for technical inspection of construction Any additional studies to be carried out Engagement of a review consultant for the review of the design drawings and to ensure compliance with the drawings approved in the stage i design and approved by the financing agency and the owner, to ensure compliance with the appropriate standards and codes and to review the construction of the facilities	Financing agency
8	Detailed design (Design Stage II)	Consultants to be selected and contracted to do the following: Advise the owner on the need for additional investigations such as soil borings and the management of the investigations Adhere to the Design Stage I and to the mitigation strategy agreed upon by the owner and the financing agency Ensure that the design principles set out in the pre-investment study are adhered to Ensure environmental standards Develop the construction program, specifications and other tender documents Select materials to be chosen in accordance with the procurement guidelines and particularly in accordance with the need for hazard mitigation Advise on the selection of building contractors in accordance with the procurement guidelines of the financing agency	Design consultants Review consultants for review of the drawings and documentation
9	Construction	Construction contractor selected in accordance with procurement guidelines of the financing agency. Contractor to provide construction program and other information as required by the construction contract. Note: Generally the contract document used for infrastructure projects is based on the draft prepared by the International Federation of Consulting Engineers (FIDIC)	Owner Construction contractor
10	Inspection	Inspection of construction to be carried out by the design consultants with reviews by the review consultant. A resident engineer and other technical staff may be employed for continuous inspection. Review consultants to carry out periodic inspections. Financing agency to carry out periodic inspections to ensure compliance with the conditions of loan approval. Payment certificates to be issued in accordance with the terms of the construction contract.	Design consultants
11	Final Inspection	Design consultant to carry out final inspection, and report to the owner and financing agency on the final costs of the project and the contractor’s compliance with the contract documents. The report should contain: As-built drawings to be prepared by the design consultants Design criteria Construction problems found and solutions to the problems List of consultants and experts employed Assessment of the contractor’s work Explanation of delays Variations and extra costs Final costs Date of start of construction and date of completion The review consultants should issue a final report to the owner on the compliance of the work with the approved technical standards and codes and with the principles established for hazard resistant construction and environmental control.	Design consultants Review consultant

Items	Costs (US dollars)
	Local costs	Foreign costs	Total
Civil works Main restoration works Temporary works	363,000 163,000	2,093,000 426,000	2,456,000 589,000
Engineering and project management	15,000	263,000	278,000
Base cost	541,000	2,782,000	3,323,000
Physical contingencies	38,000	221,000	259,000
Price contingencies	64,000	253,000	317,000
Interest during construction	-	34,000	34,000
Total	643,000	3,290,000	3,933,000
Percentage of total cost	16.3 %	83.7 %	100 %

Buildings and Infrastructure Project: A Case Study of Caribbean Infrastructure Projects that have Failed Due to the Effects of Natural Hazards

Organization of American States General Secretariat Unit for Sustainable Development and Environment

USAID-OAS Caribbean Disaster Mitigation Project April 1998

This report was prepared by Alwyn T. Wason - Principal Consultant

APPENDICES

SUMMARY RESULTS OF CASE STUDIES

REFERENCES

Buildings and Infrastructure Project:
A Case Study of Caribbean Infrastructure Projects that have Failed Due to the Effects of Natural Hazards

Organization of American States
General Secretariat
Unit for Sustainable Development and Environment

USAID-OAS Caribbean Disaster Mitigation Project
April 1998