Flood protection for a Tunisian city

! ' NUMBER 76 0139 Precis Operations Evaluation Department November 1994 Flood Protection for a Tunisian City A recent OED audit* examined the In October 1982, a flood hit Sfax, ping a maintenance unit in re- impact of a Bank-financed project de- of a severity estimated to occur gional offices of the Ministry of signed to protect the city of Sfax and every 150 years. Seventy lives Equipment, which was charged outlying areas from the devastating were lost, and more than 700 with administering flood protec- effects of frequent floods. By reducing houses were destroyed and 8,000 tion in urban areas. the risk of flood, the project stimulated damaged. Flood protection mea- development, created a feeling of in- sures were minimal. Plans for A related project, the Second creased security among residents, and more systematic protection had Urban Sewerage Project, called for eased the soil erosion common to the been under consideration since the the construction of three storm area. The project was implemented 1950s, but were not acted on for water collectors to evacuate flood very rapidly by a national ministry, lack of financing, waters from the city center. at a cost about 16 percent below appraisal estimates. The audit high- A plan for flood protection The project lights the tradeoffs common to such projects-between a quick response The Tunisian government What worked and a response that involves the ben- reacted quickly to the flood, re- eficiary community. Emphasis on questing Bank assistance for the Overall, the project was suc- rapid action precluded full participa- design and construction of flood cessful. The goal of reducing the tion by city authorities, causing the protection facilities. The Bank also risk of flood was met, although ac- project to overlook some basic issues. moved expeditiously. A project tual effectiveness, in terms of engi- This and other lessons learned should was approved in May 1983, with a neering specifications, has yet to be of value to similar emergency and total cost of $48 million. The Bank be tested. Recent heavy rainfalls, flood protection projects. was to provide $25 million, though not of flood proportions, have been evacuated with greater Background The project sought to provide ease than in the past. flood protection for Sfax and In 1983, about 313,000 residents neighboring communities at an The project enhanced urban lived in Sfax, Tunisia's second economically acceptable cost. Pro- development in Sfax. The substan- largest city. This port city lies at tection facilities were to be able to tial amount of new construction the foot of a gently sloping plane withstand a flood of catastrophic in damaged areas attests to the with sandy soil and sparse vegeta- proportions occurring once every increased confidence of develop- tion. Annual rainfall ranges from 100 years. ers, both public and private. Dike 37 mm to 650 mm, averaging about 200 mm, with torrential Works were to consist of rains striking in October. The high rehabilitating and increasing dike "aPefomance Audit Report: water table, extensive urbaniza- protection on two natural water tion, and lack of vegetation con- courses that bracket the city on the pod Po. rct tribute to intense flash floods and north and south. A semi-circular report ar c 1994. heavy erosion. "belt canal" was to be constructed Et) ecot s are at some 4 km from the city center. Backgrti undte De nts Since 1934, five major floods Plans also included minor works fro enal oc m withnstndydsoilmandgionale vegrta- have caused severe damage in to protect adjoining municipali- tion Services Centers. the city and surrounding areas. ties, and establishing and equip- protection permitted agricultural tion to the disposal of storm water caused the project to overlook land to be re-zoned for urban devel- falling on the city itself. Storm some basic issues. opment, increasing the value of the drains constructed under the urban land by up to 400 percent. The project are not sufficient to handle The Bank tends to be preoccupied project has also contributed to per- the excess water. Heavy rains are with the implementing agency at the ceptions of increased security not being drained into the canals expense of beneficiaries. The apparent among residents, particularly the and standing water remains on the gaps in cooperation between the poor. New roads constructed along roads and in residential areas, caus- government and the city of Sfax the belt canal are another visible ing health hazards among other highlight the danger of this preoc- benefit of the project. These roads problems. Sfax would not be faced cupation. Maintenance of the provide for better traffic circulation with this problem had the city au- works was vested in the appropri- (all other roads are radial) and act as thorities, as well as the central gov- ate ministry, but city authorities a downtown bypass. ement, participated more fully in should have been more involved. project design. City ownership of the project, in- The project's overall environmen- cluding responsibility for mainte- tal effects are clearly beneficial. Not The project's negative effects nance, would greatly facilitate the only is the danger of property dam- should be considered temporary as integration of flood control canals age and loss of life reduced, but their resolution is a matter of reme- into the overall drainage system, there has also been a decrease in the dial action. The real problem arises as well as helping to resolve any extensive soil erosion that accompa- when it comes to taking the action, future problems. nied each flood in the past. because there is confusion over who is responsible. The city authorities *The Bank sometimes overlooks im- Institutional development regard the canals, protective works, portant issues in "fast track" opera- was substantial. The creation of their maintenance, and any result- tions. The integration of flood con- an efficient maintenance unit for ing problems, as the government's trol works with the city's drainage flood protection works was a first in responsibility. The project did in system should have been a basic el- the Ministry of Equipment. Three fact entrust the Ministry of Equip- ement of the project's design. The similar units have been established ment with the job of maintaining issue of water conservation also in other parts of the country, mod- the flood protection works, but it is should have been given more con- eled on the Sfax maintenance unit. hard to see how the Ministry should sideration. be held responsible for the city's What failed overall drainage system. Projects of this nature need to se- cure baseline data to help determine a The project did have some nega- Lessons project's impact. Completion reports tive effects. These included accumu- claimed that land/property values lation of contaminated groundwater This emergency project was in previously flood-prone urban ar- at the northern outlet of the belt ca- processed efficiently. Project eas had increased "multifold." nal, which results in unpleasant appraisal and approval were These claims cannot be substanti- odors and mosquito breeding. The achieved quickly, implementation ated because data on property val- contamination is the legacy of an ol- was completed on schedule, and ues were not gathered at project in- ive oil factory that once operated at costs were 16 percent below ception, and cannot be determined the site, and whose waste products appraisal estimates. now because there are no property saturated the ground. Maintenance taxes in Sfax and authorities keep crews that remove rubbish from the The experience highlights the no records of property transactions. canal on a regular basis cannot cope tradeoffs in emergency projects with the problem. that may need to be made-be- In a country where water is scarce, tween a quick response and a flood protection measures should con- Further problems are created by response that involves the benefi- sider conserving flood waters instead insufficient roadside drainage sys- ciary community, and is thus likely of just evacuating them. Saline intru- tems in the city. Project design fo- to be more beneficial in the long sion limits the use of groundwater cused on evacuating flood waters run. In this case, early action pre- near Sfax. Combining flood protec- approaching the city from the hin- cluded full participation by city tion with conservation would have terland, but paid little or no atten- authorities, and a hasty design been a huge plus. OED Pr&is is produced by the Operations Evaluation Department of the World Bank to help disseminate recent evaluation findings to development professionals within and outside the World Bank. The views here are those of the Operations Evaluation staff and should not be attributed to the World Bank or its affiliated organizations. Please address comments and enquiries to the managing editor, Rachel Weaving, G-7137, World Bank, telephone 473-1719. Internet: RWEAVING@world bank.org November 1994