v. PmILE COpy Republic of Senegal THD WATER PROJECr 'TINERIM PHASE" DRINAING WATER COMPONENT Environmental Impact Study Patrick LAURAS/Jean-Luc PIGEON Consulting Engineers SETUDE 11, boulevard Pershing 75017 PARIS E9440 (M) March 1995 TABLE OF CONTENTS 1. SCOPE OF SURVEY 1 2. SUMMARY AND RECOMMENDATIONS 2 3. REGULATION AND ADMINISTRATION FRAMEWORK 5 3.1. For the management of water resources and the supply of drinking water 5 3.2. For the protection of the environment 6 4. DESCRIPTION OF THE PROJECTED INVESTMENTS 7 4.1. Resource immobilization investments 7 4.2. Investments for water transport 8 4.3. Investments concerning the functional structure of the networks 8 4.4. Investments into distribution 9 4.5. Tailoring of needs and resources 9 4.6. Technical summary of installation production conditions 11 5. STATE OF THE ENVIRONMENT AND PROJECT IMPACTS 14 5.1. General data concerning survey area 14 5.2. Underground water resources: current situation and project impacts 14 5.3. Impacts due to layout of piping and construction 19 5.4. Sanitary imnpacts connected to water and drainage 21 6. ALTERNATIVE ANALYSIS OF PROJECT 23 6.1. Absence of overall alternatives 23 6.2. Individual or technical alternatives 23 7. MEASURES OF COMPENSATION TO THE BENEFIT OF THE ENVIRONMENT 27 7.1. Compensatory measures and recommendations concerning the use of water resources mobilized by the project ........................... 27 7.2. Compensatory measures and recommendations concerning the sanitary impacts of the project 32 7.3. Compensatory measures and recommendations concerning impacts due to the piping layout 35 APPENDICES APPENDIX 1 Memos drafted at DAKAR subsequent to the 1st mission (June 1994) 39 APPENDIX 2 Memos drafted at DAKAR subsequent to the 2nd mission (February 1995) 53 APPENDIX 3 Synthetic evaluation of ground water table 70 APPENDIX 4 Maps of neighboring effects 83 APPENDIX 5 Reproduced documents 97 REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - Fax. 16 (1) 45 72 97 63 2 2.SUMMARY AND RECOMMENDATIONS The survey of the impacts of the 3rd water project upon the environment was carried in the frame of the 2 missions to DAKAR from June 13 to 24 1994 and from February 8 to 24 1995 inclusive during which the main departments concerned by the CAP VERT water supply project were encountered, with the Ministry of Hydraulics, with SONEES, with the Cayor Canal survey mission. In the field, all the installations of the installations and the layout of the new pipes have been reconnoitred entirely. 2.1. Potential impacts of the project on the underground-water resources are acceptable as long as a careful follow-up policy is applied r Impact of the new drawing off considered in the North Coastal ground-water is moderate. Potential drawing-off from the ground-water table and its piezometric evolution have been evaluated by BRGM in 1992 thanks to modelling of the ground-water table behavior and demonstrating that: water exports by evaporation are in the great majority: 300,000 m3/day in 1992 with only 70,000 m3/day drawn off; * the loss induced by the new bore holes (35,000 m3/day) will not be felt on the north-south piezometric dome which protects the North Coastal ground-water table against the invasion of salt water; the effect of the bore holes will only be felt on the GUEOUL-THIES line i.e. near the layout zone: the induced drop should be around 2.5 to 5 meters over 15 years, acceptable considering the depth of the aquifer. However, it should be pointed out that under the current climatic conditions, the resource of the North coast is not perennial and the implementation of a long-term supply project, like the Cayor canal, should make it possible to reduce the water drawn off. r On the other hand, the resources in the CAP VERT region are a justified subject of concern. A considerable drop in the level of some of the aquifers have been observed. However, these drops are not a consequence of the project but result from the current drawing off of the ground- water table on the peninsula and the limestone ground-water table of POUT and SEBIKOTANE. It will be possible to maintain this amount of drawing off in the interim phase as long as piezometric evolutions are followed up regularly and critically together with the chlorine contents, sign of a possible intrusion by salt water. The amounts of water taken for WS at CAP VERT are from ground-water which, in some cases, is barely being refilled at the present time. Therefore, we have to insist upon the fact that the supply scheme for water to CAP VERT, in the interim phase is necessarily provisional. It would be very inadvisable to count on maintaining this amount of drawing off beyond a period of 10 or 20 years. r The report recommends a number of accompanying measures for the project. To some extent, this concerns arrangements for following up the evolution of the ground water used for WS in CAP VERT: creation of 21 supplementary piezometers (already provided for in the GERSAR/BETURE draft project); REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 1 1, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 3 requirement to confrmn the organization already set up by the Hydraulics Management (Direction de l'Hydraulique)and SONEES so as to follow up on a monthly basis 88 piezometers (to which the 21 piezometers to be created must be added); need to establish a forecast frame of reference for the piezometry of the North Coastal ground-water table so as to react in the event of the real piezometry drifting with respect to the forecasts of the BRGM model under the responsibility of the water resources management and planning department; need to stipulate a regulation for checking the water drawn off by a decree applying to the water code law. and, to some extent, the recommendations and compensatory measures: * regular cleaning of the PANTHIOR dam to improve the refilling of the SEBIKOTANE ground-water table; * preliminary survey to create other water refilling retainments; establishing of a budget provision in order to be able to compensate for the drying up (or more probably the drop in the flow-rate) of several village wells near the new bore holes on the North Coast (deepening of the wells or connection to SONEES pipes) at the expense of the owner. 2.2. Impacts upon the environment and neighborhood of the installations are generally very limited r There are three potential impacts that are worth mentioning at this stage: the crossing of the PIKINE - DAGOUDANE - THIAROYE built-up area will require particular attention conceming the work on the site in order to minimize the effects of nuisance upon the populations; the construction of the MAMELLES reservoir is in a sensitive area; crossing a number of the gardens will have a temporary effect upon the "market gardening" process. REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 4 r The main recommendations made in this study concern: the need to establish particular technical arrangements on the site for contractor consulting when crossing dense urban zones is involved (sheet piled excavations, rapid covering of trenches, maintained access to homes, etc.) and for the safety of the people concerned; * the need to pay attention to how the Marnelles reservoir is inserted while complying with the natural topology of the site; * the need to provide for financial compensation in the event of temporary market garden produce loss; * the need to ensure repair of the sites for crossing the listed forest of Pout and the rewooding perimeter of Mbaw. 2.3. The sanitary impacts of the project are rather positive r The project will serve districts where inhabitants do not have running water in their homes at the present time (43,000 social connections are scheduled). This impact will be positive on the populations concerned. r However, it is certain that the water shortage will continue and that there will still be cuts, less numerous than at the present time, during the first years of the interim phase. It is important to underscore the fact that to attenuate this shortage, SONEES is committed to a water savings policy by searching for leaks and locating outdated networks, by anti-waste communication, etc. In 1994, using its own capital, it will refurbish a 40 km line (out of a total line length of 1600 km). This considerable effort have to be maintained for several years and it would be worth reinforcing it with a tertiary network renewal program forming part of the distribution section of the 3rd water project. r From another standpoint, fears about the risks of reinforced pollution caused by dumping into the sea are unjustified. r However, pollution by nitrates from the THIAROYE sand layer could lead to a problem of drainage connected to high urbanization and is liable to become increased. The current contents are already very high (150 to 300 mg/A) and therefore, it is advisable to: maintain the current device to dilute this water more before putting it into distribution (connection to ALG 1); carefully monitor the nitrate contents and their bacteriological levels; REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 5 3.REGULATION AND ADMINISTRATION FRAMEWORK Before going on to analyze the project and its impact, we have to briefly outline the regulation and administrative framework of the water and environmental sectors in Senegal. 3.1. For the management of water resources and the drinking water supply r The government is the authority as concerns water in Senegal, on the basis of the following principles: water production, transport and distribution are public and governmental services (law of July 19, 1965); hydraulic resources are part of the public domain. Their use comes under government control (law of March 4, 1981 concerning the Water Code), as is water dumping. At the present time, the Hydraulics Department at the Ministry of Hydraulics is responsible for these skills, with three different departments: one department in charge of following up the rural WS (DEM); one water resources management and planning department; one drainage division. r The production and dilution of drinking water in the urban environment has been delegated to SONEES, a national company under government authority. This "delegation" is comparable to a concession i.e. the government assets required by WS have been transferred to SONEES (law of July 5, 1983 and decree of August 1, 1983). Accordingly, investments in the drinking water supply field are made by SONEES, the owner. As far as drainage is concerned, "delegation" is more like licensing because the government is owner of the networks and equipment managed by SONEES. REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 6 r An institutional reform should be taking place in the framework of the "3rd water project". The general outline of the reform has been sketched out by consultant HSD ERNST and YOUNG: they will be the subject of negotiations in the near future. In the following, we only point out the highlights concerning the "environmental management" section of our survey, considering them as working hypotheses: * the ownership of the equipment and prime contracting of the work will be covered by a national company; * the management of the equipment, water distribution and relations with clientele will be handled by an operating company via a licensing contract; * govermmental facilities for exercising its authority and control may be improved by establishing a go-between organization between government and users (Higher Water Council) and a technical organization having the necessary personnel and a budget (Higher Water Commission). Refer to appendix 5.1 for more details. 3.2. For the protection of the environment Three documents can be mentioned with respect to this impact study: * the decree of January 1, 1950 for the protection of green areas and forestry reserves: the MBAW forestry reserve is concerned by the project; * the decree of June 27, 1978 for the protection of traditional villages (the project avoids them); * a decree regarding the protection of monuments, historical and archaeological locations. The MAMELLES location is protected in this way. REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 7 4.DESCRIPTION OF THE PROJECTED INVESTMENTS The investments projected for the interim phase to reinforce the WS of CAP VERT can be described by four main sub- systems: investments to mobilize additional resources; water transport investments; investments concerning the functional structure of the networks; investments into the distribution networks. The total cost of the investments comes to 116 billion francs to which must be added approximately 5 billion for prime contracting. Figure 1 identifies the main investments in a schematic diagram. 4.1. Resource mobilization investments During the interim phase, it should be possible to increase the resources available to CAP VERT from 191,000 to 262,000 m3/day. To do this, it is intended to raise the capacity of the NGNITH water treatment station from 39,000 to 64,000 m3/day and to sink 11 bore holes into the ground-water table of the North Coast for a capacity of 35,000 m3/day. r Bringing the NGN1TH station up to level The NGNITH works was commissioned in 1971. It was designed to handle 64,000 m3/day (some civil engineering sections were even designed for 132,000 m3/day: exhaust station). Installing several electromechanical equipment items was, however, deferred to the point that the effective capacity never exceeded 39,000 m3/day. Bringing the works up to level so that the capacity reaches 64,000 m3/day means carrying out a series of local operations: * reinforcing the pumping and discharge capacities; bringing some equipment up to level; making some major civil engineering repairs (filter bases, decanting devices); * partial automation (dosing, washing of filters, extraction of sludge from the decanter); * creation of a 3000 m3 reservoir. The cost of these investments is estimated at 2.313 billion F.CFA (SONEES estimation after devaluation). REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 8 r Creation of 11 bore holes operating at 125 to 150 m3/hour in the North Coastal ground-water table Subsequent to the BRGM study and to the water resources of the North Coastal ground-water table (results given in section "impact analysis"), it was decided to reinforce the drawing-off capacities from 27,000 to 62,000 m3/day. The holes will be located near the centerline of national highway 2, spread out over forty or so kilometers between GUEOUL and KELLE. They will be connected to the new conduit by discharge pipes having a diameter of 300 mm. They will be identical in design to the existing SONEES holes. The cost of the bore holes is estimated at 3.515 billion F.CFA (SONEES estimation after devaluation). r To accompany these investments, the piezometer network will be densified: 21 new piezometers will be installed, added to the 140 existing piezometers. The estimation amounts to 706 million F.CFA. r In the THIES region, it is also intended to install 2 supplementary bore holes to supply the city, estimated at 838 million F.CFA. 4.2. Investments for water transport The transport of the new resources from NGNITH and the North Coast area to CAP VERT will be by the partial doubling of the existing "ALG" conduit: 103 km 1000 a piping between GUEOUL and THIES and 54 km 1500 0 piping between THIES and DAKAR. In general, the new supply pipes will follow the ALG pipe. They will depart from it over a distance of several kilometers to cross inhabited or built-up areas. The cost of the 1000 0 pipe is estimated at 38.88 billion F.CFA and that of the 1500 o pipe at 39.74 billion F.CFA. 4.3. Investments concerning thefunctional structure of the networks r The project provides for the implementation of a remote management system installed at point B (arrival and dispatching point for the water resources). This will be a complete monitoring, alarm, control and measurement system working on the networks with 6 secondary plants and 49 satellites. The installation cost is evaluated at 2.9 billion F.CFA. r A large 25,000 m3 reservoir has to be installed alongside the existing reserves of MAMELLES. It will practically double the capacity of the total current reserves at DAKAR-PIKINE (35,000 m3). Costs are estimated at 2.4 billion francs CFA. REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 9 4.4. Investments of the distrihution section This concerns all the investments related to the CAP VERT distribution networks: * restructuring of the primary networks and interconnections (approximately 105 kIn); * renewal of any old leaky pipes (approximately 45 km); * extension of the distribution networks to new districts (approximately 400 km); * continuance of the "social connections" policy (approximately 43,000 new connections); * equipping of the distribution networks with meters, analyzers and pressure gauges to be incorporated into the remote management system outlined above. The distribution investments are estimated at 27 billion francs CFA of which 2.5 billion went to prime contracting. 4.5. Tailoring of needs and water resources The CAP VERT population represents approximately 2,000,000 inhabitants today. It should continue to increase at a rate of approximately 5% per annum. This extremely fast development among the DAKAR population explains the outstanding water requirement upsurge (although forecasts are based on constant requirements per subscriber): * the average theoretical requirements work out to 280,000 m3/d in 1992 i.e. the deficit regarding average requirements was around 30%; for the 1999 horizon, the average theoretical requirements will be around 380,000 m3/day; the deficit will still be around 35% in spite of recourse to new resources during the interim phase. Accordingly, the goal of the interim phase may not be to re-establish balance between needs and resources but simply to preserve the water supply situation awaiting a large-scale structuring project such as the CAYOR CANAL which should resolve the problems at DAKAR to the year 2020 at the least. This provisional nature of the drawing-off from underground resources in the interim phase is essential in analyzing impacts as proposed in the following paragraph. REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - 4.6. Technical summary of the work production conditions The layouts of the large pipelines have been defined generally by GERSAR/BETURE (modified "mission 0" reports), far more accurately than before with engineering studies concerning the detailed pre-production project: only minor changes are liable to be made to these layouts that we have checked through from end to end as part of assignments in DAKAR for the impact survey. More generally, impacts on the neighborhood are small compared to the amount of investment. This is due to the fact that the pipes are laid underground and that ground equipment is almost all located on sites that have already devoted to WS. The engineering studies in terms of the detailed re-production proLect lead to the following indications: An expected work duration of 2 years from 1996 to 1998. The implementing of the Cayor Canal is to take place between the years 2000 and 2005 and the operation of the interim phase will take between 3 and 8 years, better establishing the provisional drawing-off device regarding the underground resources (as a reference, in 2002 and according to mission 0, the deficit of the needs with respect to the resources was 4 1%). A soil survey campaign was started in July 1994 to distinguish between five types of soils: laterites, limestone marnes, plastic and sandy clays, clayey or silty sands and sands whose physical, mechanical and chemical characteristics have been determined. A detailed topographic survey was engaged on the Thies-Dakar section beginning in July 1994 with the survey of study bands of 30, 50 or 100 m and the marking out of the control axis. This topographic surveys led to the underpinnings of maps to 1 :1000th and to lengthwise detailed profiles down to 1 :100th in elevation in urban areas so as to define the constraints of the layout accurately. From Thies to Dakar, the layout will run parallel to the existing pipe (ALGI) except in two places: - the descent from the Thies plateau to allow for the right of way of the Cayor canal, - the crossing of the Thiaroye-Pikine urban area, characterized by fast and uncontrolled development of urbanization and by a group of constructions that it would be difficult to penetrate. The layout in the Thiaroye-Pikine urban area was retained, after land surveys and acknowledgement of restructuring projects provided for on the site. Accordingly, the layout follows that of the Bonna pipe in the crossing of the Mbaw wooded perimeter, runs along that perimeter via the North, follows the future layout of the motorway to the South in the unconstructed area (North of the police station), crosses a dense dwelling area, following the streets so that the machinery can pass without the homes being demolished, runs along the North motorway layout in the unbuilt parts of the Faidherbe camp, follows the main tracks of the Guinaw Rails to avoid any damage to the existing constructions and meets back on the layout of the existing ALG to the North at the intersection with the SNCS track at Pikine. At this level, two homes will have to be destroyed as they fall within the right of way of the existing pipeline. Between Thies and Dakar, the technical characteristics of the works will take into consideration passages of infrastructures (SNCS track) to enable one of the two ALG pipes to be used in the event of maintenance or repairs while maintaining most of the flow, and integrate the future reinforcing constraints in the Cayor Canal phase. Finally, the part of the development chosen gives privilege to the visual and landscaped impact while using a totally buried layout between Thies and Dakar. The production construction conditions are as follows: - minimum right of way of 20 m during work on site outside ofthe urban areas (10 to 12 m after construction), - excavation material placed on the existing pipe but limited to a height of 3 m, or evacuated for sites in the urban areas, - the possibility of setting slight deviation angles on the pipes to avoid the main baobab trees in the line of the layout, - maintained water services during the work, - sheet piled excavations for the crossing of urban areas, - work carried out in the dry season in areas sensitive to erosion (e.g.: the Thies plateau). Acknowledgement of pipe laying constraints will require the following of the contractor: - servicing and maintenance of all the tracks to be used, - setting up signposting and providing the necessary information, - maintaining all the passages providing access to the land lots and the circulation of livestock, - minimizing impact upon the surrounding crops by avoiding destruction of hedges and tall standing trees, - the setting-up of all necessary arrangements to prevent the run-off of site water, - reduction to 15 m of the right of way in crossing the Mbaw perimeter. 5.STATE OF THE ENVIRONMENT AND PROJECT IMPACTS Analyses of the initial state of the environment and the impacts of the project have been grouped together here to make for easier reading regarding problems resulting from the project. 5.1. General data concerning the survey area The population of the DAKAR built-up area is estimated at 1.58 million inhabitants in 1993 and it is expected to reach 3 million inhabitants in 2010 (i.e. 25% of the country's total population) with an annual growth rate of 5 to 6%/annum (directorate of forecasts and statistics). The climate is sub-desert tropical with a dry season and a wet season from July to September when 90% of rainfall occurs. The rainfall in the project area is 400 mm/annum with heavy showers sometimes exceeding 150 mm in 24 hours and above all, 50 mm in one hour. The mean annual temperature between 1955 and 1989 was approximately 24°C with a maximum in October (31.8°C) and a minimum in January (15.3°C). The drinking water supply in the survey zone is by the SONEES system drawing water from the ground water at CAP VERT (cf. appendix 3) with a supply pipe from the Guiers lake. Volumes are 191,000 m3/d; 60% is distributed by domestic connections and 40% by fountain points. Drainage in the DAKAR area is essentially by individual systems concerning 870,000 inhabitants. There is a drainage network in DAKAR for 646,000 inhabitants but only 454,000 are connected to it (a connection factor of 70%). The total waste water flow is estimated at 180,000 m3/day of which 37% (66,000 m3/day) is collected in the drainage system. 5.2. Underground-water resources: current system and impacts ofprojects The project is based upon an increase in the amount of water drawn off from the North Coastal ground-water table, from 27,000 to 62,000 m3/day; in addition to the resources currently being exploited. This drawing-off from underground resources will have extensive potential impact, by far the most significant of the project. That is why the data and the surveys regarding the potential of these resources and their evolution have been examined with the greatest of care. Accordingly, a synthetic data sheet per aquifer has been established from the database compiled during the mission (appendix 3). r THE NORTH COASTAL GROUND-WATER TABLE The amount of water now drawn off the North Coastal ground-water table has been re-evaluated by BRGM in 1992 subsequent to a field survey. The survey also made it possible to update the hydrodynamic data and work out the current ground-water table inlet/output balance. The results are as follows: * The most important demands result from evaporation which has barely been compensated for by rainfall since 3 1975. For the year 1991, it is estimated at 300,000 m /day (depending on sectors, corresponding to an evaporated water head included between 3 and 45 mm/year). To this evaporation balance we must add the drawn off water, amounting to 70,000 m3/day in 1991 (see detail in appendix 3). * Accordingly, the level of the ground-water table has dropped 50 cm per year between 1975 and 1991 in the piezometric dome region. On the RN2 line, it dropped 10 cm per year outside of the perimeters of influence of the SONEES bore holes. It is expected that over-exploitation amounting to 35,000 m3/day will occur during the interim phase. The BRGM survey has made it possible to evaluate the evolution of the ground-water table piezometry up to the year 2010, thanks to a hydrodynamic model which was appropriately set on the 1975-1991 period (1052 grid sections measuring 3 km times 3 km). In appendix 5.2 we have added conclusions concemning the modelling and the two more significant schematic charts: - the first refers to the expected drops in the ground-water table without over-exploitation (this is the "reference state" without a project) for the horizon 2010; - the second gives the expected drops, taking the project into consideration, for the 2010 horizon. Modelling demonstrates: - that the North Coastal ground-water table will drop, in any case, from 5 to 15 m by the 2010 horizon, even without the project (assuming that the weather conditions of the last 15 years prevail); - that the supplemental drops caused by new drawing off will be negligible regarding the North-South piezometric dome which it is essential to maintain to protect the ground- water table from the influx of salt water; - that the additional drops induced by new drawing-off on the GUEOUL-KEBEMER line will be around 2.5 to 5 m; - that the drop induced in the new bore holes themselves will be 14 m. Finally, BRGM observes that the new drawing-off will not induce any risk of the intrusion of the salt water in the North (LOUGA region). There are three conclusions resulting from the above: 1/ The impact of the new bore holes has to be put into proportion regarding the depletion of the resources that are not renewed in the current context of the rainfall deficit: on the scale of the North Coastal ground- water table, the new drawing-off only represents 10% of the total evaporation recovery, approximately 300,000 m 3/day. The drops are negligible at 20 km from the bore holes. 2/ The local impact of the holes could be considerable. In particular, it is probable that some wells or village holes near the new bore holes will dry up. However, this also has to be put into proportion: at the present time, SONEES is supplying water to 240 villages near the RN2 highway via taps from the existing WS pipe. The number of wells and bore holes along RN2 that are in activity is unknown but it is probably not very high. Whatever the case, this problem has to be taken into consideration and we will refer to it when dealing with compensatory measures. 3/ We must insist upon the fact that under the current weather conditions, the North Coastal resource is not perennial and the implementation of a long-term water supply project like the Cayor Canal should reduce the amount of water drawn off. r THE CAP VERT REGION GROUND-WATER TABLE The first remark to be made is that we should take into consideration the fact that the evolution of the ground-water table in the CAP VERT region is not in actual fact an impact of the project because bore-holes already exist there and there will be no more in the future. However, impacts of the DAKAR WS system are too considerable to omit describing the "environmental state" of this ground-water table. In the following, we refer only to the more salient facts (see the synthetic data sheets per aquifer in appendix 3). The infrabasaltic sandy ground-water table The infrabasaltic sandy ground-water table under the Cap Vert peninsula has been over-exploited at certain times (between 1970 and 1985). This has resulted in some encroachment of salt water. The situation has been stabilized for the last ten years or so. Maintaining the current level of drawing-off (19,000 m3/day) is possible as long as the rainfall is sufficient because the ground-water supply appears to be connected to it. Section 7 deals with the question of following up piezometry and salinity which are essential because of this risk of salt water encroachment. The THIAROYE Quaternary sandy ground-water table The operation of this layer has been reduced and the actual level has stabilized generally since 1985. The objective of drawing off 8,000 m3/day during the interim phase, corresponding to the volumes drawn off over the last few years is therefore not a problem regarding the resource (as long as the rainfall is sufficient). However, we should observe that the quality of the water has degraded over the last few years with a considerable increase in nitrate content: contents close to 300 mg/l have been measured while the WHO recommends 50 mg/l as a limit for drinking water supply. This pollution results from the infiltration of waste water from PIKINE-THIAROYE, a built-up area covering the entire peninsula area where there is no drainage. The SEBIKOTANE Palaeocene limestone ground-water table First and foremost, the ground-water we are about to refer to now is a single but complex aquifer system (SEBIKOTANE and POUT limestones and POUT Maestrichtian). The SEBIKOTANE ground-water was exploited enormously in the beginning of the 60's causing salt water to invade its southern part. Drawing-off was cut down to some extent and now amounts from 25,000 to 27,000 m 3/day. It appears that the encroachment of the salt water is continuing, quite slowly, in the southernmost tip. The chloride contents of some piezometers have increased considerably since 1985 (increasing from 35 to 300 mg/l). Therefore, the ground-water is still slightly over-exploited and maintaining the current level of drawing off is only possible with close supervision and as part of an overall project for the DAKAR WS in a relatively short time. The POUT Palaeocene limestone ground-water table The ground-water here is in a similar state to the SEBIKOTANE system: over-exploitation during the 60's leading to salt water invasion; at the present time, the drop is 60 cm per annum; chloride has increased in the south (120 to 275 mg/I between 1985 and 1993). For the interim phase it is intended to draw off 20,000 m3/day in all. This is a little more than in 1992 but nevertheless down with respect to the years before 1991: the commissioning of the emergency phase bore holes in the Maestrichtian areas was accompanied by a deliberate reduction in the amount of water drawn off the Pout limestone layer because of the connection between the two aquifers. In the same way as for the SEBIKOTANE ground-water, the scheduled volume is only acceptable if piezometry and salinity are monitored very carefully and if the perspective of an overall project for the DAKAR WS is confirmned within a reasonable period of time. The POUT Maestrichtian ground-water table The Cap Vert WS urgent phase involves a major increase in the amount of water drawn off by SONEES in this deep- lying aquifer: from 37,000 m3/day in 1991 to 74,000 m3/day in 1994. During the interim phase, the drawing-off from the ground water will remain at this level of 74,000 m3/day, reached during the emergency phase. Other amounts drawn off will be considerable. They were estimated at 50,000 m3/day (in 1985). Therefore, total drawing-off will be around 120,000 m3/day (see appendix 3). With this amount of drawing off, the ground-water will drop by 1 m to 1.5 m per annum, acceptable without difficulty for 10 or 20 years considering the great depth of the aquifer. To summarize the above, there is more to be concerned about regarding the ground-water near DAKAR and the Pout and Sebikotane limestone ground-water than the increase in exploitation of the Maestrichtian and North Coastal ground-water. However, we must bear in mind that the water being drawn off here will come from quasi-fossil resources (not renewed in the current context of rainfall deficit) which will not settle the problem of the long-term water supply for DAKAR. r IMPACT OF DRAWING-OFF IN THE GUIERS LAKE The MANANTALI and DIAMA dams will ensure that the Guiers lake is supplied with flow-rates of around 40 to 50 m3/s, even during the dry season. The water drawn off at NGNITH, 0.7 m3/s is negligible. However, we have to face the facts conceming the ecological problems that could confront the Guiers lake in the next few years: increase in the pesticide factor, loss of capacity due to sedimentation. However, none of these would be the result of drawing off water from NGNITH. 5.3. Impacts due to the layout of the pipe and the constructions: Appendix 4 is a series of maps to a scale of 1:50,000 showing the main points of impact that we describe below: r On the GUEOUL/TMES line, there is very little impact if we take care to place the pipe on the western side of the existing ALG pipe and to skirt around ten or so hamlets or villages so that we do not interfere with the fencing. In particular these are: N'DIAYE TIORO, BEUD FORAGE, DAKAR NGOGNE, PALMEO, GUEOUL NDANDE (sacred tree), etc. Almost everywhere, in fact, the ground is used for traditional millet and manioc crops. There are also two large swamp crossings where the pipe must be laid overhead on concrete posts (and where the landscaping impact is of no consequence). r On the THIES-POINT K section, the sensitive areas through which the layout passes comprise market garden areas and orchards (Niayes area), the listed forest at Pout and many talwegs with temporary flow conditions. * In the crop-growing areas, consisting of traditional family market gardens and large market gardening ventures, the passage of the pipe will lead to a temporary loss of operation. * In the Pofut forest, the 8.5 km of layout will lead to the elimination of vegetal elements (hedges) but, in theory, no tree felling. In the talwegs, passage of the pipe will lead to the cut-off of flow and risks of erosion. r On the POINT K-POINT B section, we note three types of impact, in increasing order of importance: the crossing of the rewooded perimeter of MBAO following the existing layout of the BONNA, the crossing of many gardens, more often than not used for vegetable farming. The impact will be transient; the crossing of dense built-up area between THIAROYE and PIKHNE. The different solutions we have considered are analyzed in paragraph 6.2 below. r The construction of the MAMELLES reservoir (25,000 m3) is the most significant individual impact of the project. The Mamelles site is listed as regards the protection of historical and archaeological areas but this classification does not correspond to the precise criteria of the environment: the site is agreeable but nothing exceptional and the environment will soon be urbanized. In paragraph 6.1, we will see that there is no possible alternative to install a large-volume reservoir (which is indispensable for the continuity of distribution in the MAMELLES sector (i.e. the entire North of the peninsula) and for the safe distribution to DAKAR center. Careful insertion of the reservoir is recommended as regards compensatory measures (see paragraph 7.3). r Finally, we have to deal with problems of impacts connected to the site. More particularly, these include: the hindrance to nearby residents in all the streets where pipes will be replaced (distribution section). This is temporary and conventional for all road works; risks of accidents involving people living near the villages and in the built-up areas, connected to the opening of a trench approximately 3 m deep in highly frequented areas (see compensatory measures). 5.4. Sanitary aspects connected to water and drainage Increasing the volumes available for drinking water supply will lead to the possibility of distributing water to a wider population and the probable increase of individual consumption. This will have sanitary impacts and consequences upon the purification of waste water. r SANITARY IMPACTS The first sanitary impact of the project is positive: it will extend service to the home thus considerably improving the sanitary state of the dwellings. The "social connections" lot of the "distribution section" is essential on this point: it is intended to create 43,000 connections, meaning that the total number will be increased by around 30% from 122,000 to 165,000 subscribers (from 1990 to 1993, 41,000 social connections will already have been made - see appendix 5.6). Note that the extension of the water networks is not a factor of incitation to urban immigration because the new districts are developing well ahead of the supply of water services. In areas where there are drainage networks, there could be a considerable sanitary effect related to the increased flows to be evacuated. Will the hydraulic capacity of the used water networks be sufficient and are there not risks of waste water overflows? According to all the people questioned on the subject, the current problem is rather one of flow insufficiency in the networks and the considerable depots resulting therefrom. Accordingly, there is no risk of overflow of used waters in the interim phase. r DRAINAGE AND POLLUTION OF THE NATURAL ENVIRONMENT Several people raised the problems of the effects of the increased volumes of water consumed regarding the pollution of the natural environment. Distinctions must be made between several different cases: In areas where there is no drainage, waste water is often evacuated by infiltration into the soil by means of pits or infiltration wells (this is the case in the entire PIKINE - THIAROYE - GUEDIAWAYE built-in area where the soil is sandy). The fact that running water is available near the home means that people consume more, probably favoring the phenomenon of infiltration and accelerating the transfer of pollutants towards the underlying aquifers. This mechanism is a problem regarding the drinkability of water in the THIAROYE sandy ground-water used by SONEES at a rate of 8,000 m3/day: the nitrates have already reached levels of around 150 to 300 mg/liter and it is probable that the increase will continue because of the urban planning itself and subsequently as a consequence of the acceleration in infiltrations (see recommendations in paragraph 7.2). Conversely, the reinforced drinking water supply has barely any immediately incidence upon the pollution flows dumped into the sea via the used water networks. The pollutants dumped by a population depend relatively little upon the amount of water consumed. The fact that water cuts are limited does not really modify the lifestyle of a household and its "production" of polluting matter: dumping simply becomes less concentrated. Accordingly, the polluting impact of direct dumping into the sea will not be modified. However, it could have an effect upon the treatment stations: their hydraulic load could increase because of the dilution of the waste meaning that it would be necessary to limit the amount of pollution accepted for treatment. However, in our opinion, this effect will be very small during the interim phase (there is still a water shortage and drain- off will therefore be highly concentrated) and in any case will only concern pollution currently being treated (the main station, CAMBERENE, has a capacity of 100,000 equivalent inhabitants only while the population provided with drainage is estimated at 470,000 inhabitants). Obviously, it is indispensable that these drainage schemes, currently under study, take into consideration the future water volumes. On this point, the study of the DAKAR drainage scheme carried out by JICA, is a considerable anticipation because the retained volumes correspond to the "CAYOR CANAL" phase. Nevertheless, we should bear in mind that this survey does not mention the need that it will be eventually be necessary to increase the hydraulic capacity of the CAMBERENE treatment station: it has been designed to handle highly concentrated waste (9,600 m3/day for 6000 kg of DBO5). When the effluent is more diluted, because of the increased water consumption, the biological capacity may no longer be possible because of hydraulic insufficiency. 6.ANALYSIS OF ALTERNATIVES There is no general alternative for the water supply to Cap Vert in the short term but there are a number of technical alternatives concerning the definition of the installations. 61. The absence of an overall alternative The idea of doubling the ALG pipe dates back to the beginning of the 80's. The feasibility study was by RHEIN- RHUR but the project did not go through because the CAYOR CANAL at the time, was seen to be a far more "structuring" alternative by the Senegalese authorities. The difficulty and lengthy nature of the CAYOR CANAL surveys soon came to the forefront, so that in 1985, the studies were restarted for a transient phase of water supply to Cap Vert while the perspective of creating the CAYOR CANAL was maintained for the medium term. This strategy is still current: a major long-term project is essential. But the production delays justify, considering the situation of the WS (in some districts, cut-offs are repetitive and last from morning to midnight), an "interim phase" to be implemented more quickly. Therefore, the CAYOR CANAL should not be analyzed as an alternative to the interim phase project but as a medium-term perspective. What is more, there is no overall alternative making it possible to increase the DAKAR water resources by 60,000 m3/day in a three-year deadline. 62. Individual or technical alternatives However, there are some technical alternatives concerning the definition of the installations, mainly already analyzed by BETURE as part of the "modified mission 0": r The first major alternative concerns the over-exploitation of the North Coastal ground-water. The modelled survey of the ground-water by BRGM in 1992 considered three over-exploitation hypotheses with respect to the current situation: +25,000 m3/day, +35,000 m3/day, and +50,000 m3/day. Although the level of 50,000 m3/day could have been considered, the Senegalese authorities wisely chose a 35,000 m3/day level. The impact of this drawing-off is submitted in paragraph 5 above. 1 I.SCOPE OF SURVEY The "3rd water project" is an overall development project for the water sector on the CAP VERT peninsula. It has been divided into two major phases, the first "urgent" phase which, in 1992, already made it possible to increase the available resources by 40,000 m3 per day thanks to 10 new bore holes in the POUT region, into the Maestrichtian deep table, and a "interim" phase which should be started during the forthcoming months. This interim phase will include: * considerable investments to reinforce the CAP VERT drinking water supply (estimated at 116 billion francs CFA - see figure 1); an institutional reform of the water sector; complementary surveys concerning the water sector (reuse of water by market gardeners, drainage studies, etc.); if necessary, complementary investments such as, for instance, the drainage of Rufisque. This survey is designed to evaluate the impacts of the investments provided for in the water supply to Cap Vert and the environmental follow-up measures accompanying the project, while taking into consideration the perspectives of institutional reform. A number of investments which originally formed part of the "3rd water project" program are already underway or will be financed in another framework: the "WS Small Coast" project is underway, with CFD financing; the "WS eight centers" project is underway with JICA financing; the "WS six river centers" project is being surveyed (with financing by KFW); the "WS eleven towns" project should also be financed by KFW; the feasibility study is underway. Therefore, these projects were not examined during this impact survey. REPUBLIC OF SENEGAL Environmental Impact Study Patrick LAURAS / Jean Luc PIGEON THIRD WATER PROJECT SETUDE - 11, bd Pershing 75017 PARIS INTERIM PHASE Tel. 16 (1) 45 72 97 60 - r The second alternative we have considered concerned the capacity of the GUEOUL-THIES line. Three types of reinforcement were considered: reinforcements by overpressure, by the creation of a new 900 6 pipe or by creating a new 1000 6 pipe. GERSAR/BETURE made relatively tricky economical simulations, depending on the date on which the CAYOR CANAL was commissioned and on whether or not the North Coastal ground-water table continued to be used after the commissioning of the canal. These simulations demonstrated that if we apply the current 8% discounting factor, the only hypothesis favorable to doubling by a 900 6 pipe is that in which the CAYOR CANAL is put into service in 1999 and where over-exploitation of the North Coastal table stops immediately. The hypothesis of over-pressure only holds together if the CAYOR CANAL is commissioned as of 1997 (resulting from particularly penalizing energy costs). Obviously, this hypothesis is unrealistic. Therefore, the real choice falls between doubling the 900 6 pipe or the 1000 6 pipe. The difference in investment costs is 6 billion francs CFA (after devaluation: 37 billion against 43 billion). This additional expense appears to be justified considering the risk we ran that the CAYOR CANAL is only commissioned after the year 2000. In terms of impact upon the environment, the alternative between a 1000 6 and a 900 O pipe is insignificant. r Similarly, alternatives concerning the dimensioning of the THIES-POINT B pipe have been considered: 1200 6, 1500 6 or 1700 0. Here again, economic analysis made it possible to opt for the 1500 6. T his choice has no fundamental effect upon the impact of the project: in all three cases, extensive construction work and the laying of pipes is involved, requiring a right of way of between 8 and 10 m width. r Alternative layouts of the POINT K-POINT B section have also been examined. This section has to pass through 2 km of dense built-up area in which the right of way will infringe upon land occupied by existing houses or buildings. Two reasonable solutions have been identified: along the national highway parallel to the existing ALG, or along the right of way of the future motorway (seefigure 2). * along the national highway, a good fifty or so buildings and dwellings of satisfactory quality will have to be moved. In addition, the impact of the working period on the construction site would interfere with many activities taking place along the road and where traffic is considerable; that is why this layout was not retained. - along the right of way of the future motorway, the layout suggested by GERSAR/BETURE, two sections of approximately 800 m will pass through a densely inhabited area, referred to as uncontrolled habitat, where summary constructions exist. This solution definitely has a more moderate impact because it is not necessary to destroy the dwellings in spite of the narrowness of the streets. Furthermore, it is important to point out that this "uncontrolled" area will be the subject of restructuration in the more or less short term with respect to urban planning (it is impossible to install a drainage network in this area) and streets (by creating and main and secondary streets, right of way for the motorway, etc.). In the crossing of the first dense urban area at Pikine, there is an alternate layout mentioned above with a restructuring project coordinated by GTZ which, in the initial phase of work, provides for the construction of a 20 m right of way road that could follow the ALG pipe. This project, initially scheduled for 1997, could take place before this date if the financing of the first emergency phase is acquired. As things now stand, a complete real estate survey and topographical measurements are complete on the layout of the 20 m way. The urban planning directorate recommends coordination between the two projects in this zone. r Finally, BETURE examined various alternatives for locating the MAMELLES reservoir. This reflection took into consideration the fact that in the fnal term, 75,000 m3 of water will have to be stocked. First and foremost, there is no other site near DAKAR whose altitude would be sufficient for a gravity drain system: because of the volumes to be stored, there is no question of using raised reservoirs other than installing large pumping systems to repressurize the water for distribution, with considerable energy costs (estimated at 15 F CFA by BETURE). This means there is no alternative location. With the location selected, the choice of set-up proposed by BETURE was governed by technical and economic considerations. The choice, in fact has little effect upon the impact of the structure which will be essentially connected to architectural insertion and to possible landscaping (see the compensatory measures below). 7. COMPENSATORY MEASUREMENTS IN FAVOR OF THE ENVIRONMENT The compensatory measures and recommendations concern the use of the water resources mobilized by the project regarding the sanitary aspects and impacts due to the layout of the pipeline. 7.1. COMPENSATORY MEASURES CONCERNING THE EXPLOITATION OF WATER RESOURCES MOBILIZED BY THE PROJECT Obviously, it is impossible to propose compensatory measures (strictly speaking) for the exploitation of underground resources. In this area, technical supervision and management of resources has to be a priority. Accordingly, the impacts upon natural resources, as indicated previously (chapter 5) will lead to a series of compensatory measures first aimed at reinforcing the follow-up of the evolution of these resources. 7.1.1. The need to confirm the current underground resource management organization and a North coastfollow-up framework The water resource management and planning department (Ministry of Hydraulics) and SONEES decided to pool their means at the beginning of 1993 to ensure the monthly follow-up of 78 piezometers in the Cap Vert region and 10 piezometers on the Northem coast. The measurements are made each month by SONEES personnel, in some cases accompanied by a representative from the water resources department. The announced goal is to publish six-monthly balance sheets - an example is given in appendix 5.3. These arrangements must be followed in practice and the balance sheets must concern all the aquifer systems concerned by the CAP VERT WS. They will be readjusted to the framework resulting from the institutional reform: the operating company must, for instance, be engaged contractually in carrying out ground surveys with annual checking; but a body closer to the government (central water commission?) will be entrusted with the task of interpreting them and supplying the related orders. Finally, at this point, note that 21 new piezometers will be created during the interim phase (the distribution per aquifer is specifed in appendices 3.1 to 3.6). These piezometers will fill in some of the gaps that are evident in the current system. Naturally enough, it is indispensable for them to be integrated into the piezometer network followed up on a monthly basis by SONEES and the Ministry of Hydraulics. Furthermore, the Directorate of Hydraulics raised the specific problem of the advanced risks of the saline area to the North of LOUGA. Although the BRGM model does not show this risk, much attention will be necessary in carrying out the piezometric readings in this region. Implementation conditions: The BGR report concerning the evaluation of the ground water exploitation in the Pofit/Sebikotane region (October 1994) and SONEES observations made in this document, stipulate the operational provisions for following up the resources as mobilized by the interim phase and the aquifers concerned by the Cap Vert water supply: to cover any short-comings in the piezometer network data, the operating company will have to carry out a topographical levelling of the drill holes and piezometers whose elevation is uncertain (to be carried out during the first half of 1995). a construction localizing map will be drawn up, integrating the piezometers followed by the central laboratory, the operating company must complete the rehabilitation of all the drill hole guide tubes feeding Dakar to ensure that level measurements are reliable (scheduled for the first half of 1995), the establishing of a data bank, recommended both by BGR and SETUDE for production, piezometry and hydrochemistry of the ground water levels will be carried out using data processing. An evaluation report on the resources will be submitted annually. Further, the SONEES production department will establish a weekly balance (follow-up. chart) of the volumes produced and the consumption requirements; these data will be compared against the production objectives determined for the interim phase and it will be possible to thus identify deviations with respect to forecasts. the installation of 21 additional piezometers linked with the project will take place before the new bore holes are put into operation. as far as the North coast is concerned, more particularly, the BRGM hydrogeological model, set appropriately, will be used for establishing a reference basis for permissible drops from year to year. It is fundamentally important for piezometric follow-up to be interpreted according to a reasonable reference basis. Accordingly, it should be recommended that modelling to follow-up the evolution of the ground water will be made available to the Ministry of Hydraulics Water management department (manual transmission of procedures for getting to grips with the software). This work will be accomplished as soon as the final locations of the bore holes and piezometers are known. It will be in the form of a grid, by year and by piezometer, (with ranges if necessary according to rainfall). It should require between fifteen days and one month of engineer time. Z..2. Increased care concerning Cap Vert ground-water tables As far as the Cap Vert region ground-water tables are concerned, it will be necessary to reduce, without hesitation, the water drawn off these aquifers which are showing signs of piezometric weakness or the invasion of salt water; compensation for this reduction could come from reinforced drawing off in the Maestrichtian ground-water tables. Implementation conditions: Reinforced drawing-off from the Maestricht resources, recommended both by BETURE/GERSAR and BGR, should be confrmed with the construction of a new 5000 m3/day bore hole at Pout (localizing survey underway). It has been observed that the piezometric levels are generally lower after three years exploitation during the urgent phase than provided for in the ARLAB model, it is also intended to study the complementary exploitation conditions because of the acute water production deficit at Dakar. To do this, the upgrading of the hydrodynamic modelling system can be envisioned, using the ARLAB model or that of the BRGM (resetting of the model estimated to cost 60 ME CFA). 7.1.3. Regulation of drawing off to be confirmed From another point of view, it is important to note that the current lack of knowledge about the village wells and bore holes and the volumes drawn off (this situation is a departure from the satisfactory knowledge of the ground-water piezometry) means that only SONEES, the large agricultural and industrial centers who are large consumers, supply the administration records of the volumes they actually draw off. The law of March 4, 1981 provides for the principle of prior authorization but the fact that there is no application decree means that drawing off prior to this date is unknown, in the same was as the volumes actually exploited. This short- coming is evident when an endeavor is made to optimize the establishment of new bore holes. Implementation conditions: Three decrees are proposed to be implemented at the end of 1995: 1. conceming the prior drawing-off authorizations (for all pumping exceeding 5 m3/h); 2. concerning the regulation of dumping into the natural environment; 3. concerning the conditions of supervising the quality of water (water policy) and the provisions applied for protecting the resources. 7.1.4 Compensation for the drying up of wells near the new bore holes The installation of eleven new bore holes on the North coast will lead to additional use of the GUEOUL-KEBEMER axis. In any case, a provision designed to fnance the deepening of wells or bore holes which dry out after the commissioning of new bore holes will be included in the project investment budgets as part of the compensatory measures (or to fnance connection to the SONEES system). Implementation conditions: The localizing of the new bore holes was handled by BRGM who installed most of the installations in the area and is therefore well aware of the different constraints (BRGM 1992 survey into the re-evaluation of the North coast water resource). Accordingly, the installation of new systems drawing water from a depth of 90 m was carried out on the basis of a bore hole survey managed by DEM (rural hydraulics) and the estimation of the village requirements (12 villages in the Louga region). The influence of the piezometric drop upon sources located nearby is estimated as eventually reaching 5 m between Gueoul and Kebemer. If existing bore holes have been taken into consideration, in the longer term, there is still a possibility of the traditional wells located nearby at an average depth estimated at 40 m (BRGM 1978 data) drying out. Note that the average operating flow-rate of a well is estimated at 7 m3/day and that village requirements are summarily evaluated at 3300 m3/d in the Ndande, Mbedienne and Sagatta districts, those being the regions closest to the new capture points (source: field survey - BRGM - 1992). Moreover, the village wells have not yet been surveyed and in general, the villages located along the ALG conduit will probably connect to it. Accordingly, there is no way of accurately defining the consequences of the new bore holes on the village wells in particular as this impact will not make itself felt immediately. To set a framework for the compensatory measures that could be taken, it is proposed to define an initial risk zone immediately near the bore holes and within a 2 km radius (this distance corresponds to the necessary distance between two bore holes) where connection to the SONEES system would be applied in the maximum risk scenario (wells drying out). From this initial zone, and up to 10 km around the bore holes, it could be possible to finance the deepening of the wells that dry up or whose flow-rate is found to be insufficient to cope with requirements. Application of these measures would be at the expense of the owner. 7.1.5. Periodic drainage of the Panthior dam (recommendation) The PANTHIOR dam should be drained every two or three years to improve the refilling of the SEBIKOTANE ground water table. The studies consulted on this subject have a generally favorable opinion or include reservations regarding its efficiency (refilling is local and temporary). In case of doubt, drainage is advisable because the stakes are considerable (incoming saline water in the South of the ground water from the Sebikotane Palaeocene limestones). The cost of this measure is estimated at 80 MF CFA (source: Directorate of Hydraulics - June 1994). 7.1.6. Preliminary studyfor the creation of other retentions to refjll the ground water (recommendation) It would also be worthwhile initiating a feasibility study into the creation of dams of the same type on the "small contributing basins" in the POUT and SEBIKOTANE regions. Some sites have already been considered as suitable by the Ministry of Hydraulics water resources management department (the POUT and SOMONE marsh and quarry areas). 7.1.7Z Supervision of the GUIERS lake (recommendation) The ecological problems of the GUIERS lake, essentially, are having an effect upon the possibility of producing drinking water for DAKAR: the phenomenon of eutrophization, at times, can modify the efficiency of the treatment lines because of the proliferation of organic matter; micropollutants are difficult to process and the admissible concentrations in drinking water are very low. It is possible to monitor these two problems: SONEES now has the means of carrying out micropollutant analyses. ORSTOM has begun to pay attention to this phenomenon of eutrophization. This monitoring work is important concerning the long-term stakes for the DAKAR drinking water supply. However, they are only of interest if they can arouse the implementation of conservation policies involving all the partners concerned by the water resource of the GUIERS lake. The recent example of the founding of a regional committee in Saint Louis, subsequent to the pollution of the SENEGAL river, associating users together sets down the way (this did not actually concern the GUIERS lake but the example would fit the case). This should be taken into consideration when responsibilities are attributed to organizations derived from institutional reform: will the central water commission have the means of following up these questions? Could the higher water committee organize deliberation between users for problems of a local nature? At this point, we must also refer to two other problems raised by certain contacts: sedimentation and salinization. The first is worrying the Hydraulics Management (Direction de l'Hydraulique), in particular with respect to the storage capacity of the lake. However, the age of the lake (20,000 years) makes this risk particularly relative, even in the medium term (which does not mean that the study of sedimentation phenomena in the lake is of no great interest; however, it is not a concern for the DAKAR DWS). Finally, salinity was an extremely serious problem before the MANANTALI and DIAMA dams were commissioned. For the last two years, it has been reduced with chlorine levels of less than 100 mg/l. Obviously, they should be monitored because the draining of the salty earth from the region will always generate sodium chloride. But they are not a source of concern for the DWS considering the hydraulic management permitted by these dams. 7.2. COMPENSATORY MEASURES AND RECOMMENDATIONS CONCERNING THE SANITARY IMPACTS OF THE PROJECT The sanitary impacts of the project are more generally positive: it will serve districts where inhabitants do not have any water in the home at the present time (43,000 social connections are expected). This impact is positive for the populations concerned. However, the water shortage will continue and there will always be cut-offs, less numerous than now, during the first years of the interim phase. Therefore, we have to continue the water savings policy undertaken by SONEES. Furthermore, pollution of the THIAROYE sand ground water by nitrates is liable to increase because there are no drainage infrastructures and the population numbers are increasing. Accordingly, it would be a good idea to make recommendations concerning the follow-up of the distributed water standards. 7.21. Continuation of the water savings policy SONEES has undertaken a water savings policy the major outlines of which are as follows: research and repair of leakage (SONEES employs a permanent detection team with a correlator); the identification of outdated networks: 36 sector meters have just been installed and a team is now constantly working on the statistical interpretation of subscriber meter readings and the reading of these sector meters; the renewal of 40 km of piping and 4400 connections in 1994 (so far the renewed length has not exceeded 10 to 15 kIn). A budget of 470,000,000 F.CFA was drawn from the capital of the SONEES fmancial plan in 1994. These figures are to be compared with the total length of the networks: 1600 kin, and the number of connections, 122,000 in 1993 (see appendix 5.4); communication campaigns aimed at saving water and avoiding waste; a water economy policy by the privatizing of fountain outlets (see appendlx 5.5). It is evident that the water shortage is such that a deficit will continue throughout the interim phase: water savings measures will therefore attenuate the deficit but not reduce the amounts drawn off underground resources. However, they are part of the essential management measures related to the project. Accordingly, we can make four recommendations on this level: SONEES or the future operating company must maintain a renewal budget for at least 40 km and 4000 connections per annum for several years until the yields of the networks reach 80-85% at the least (budget of 470 MF CFA inscribed in 1994). In addition, it would be worth incorporating into the 3rd project an exceptional requalification program for the distribution networks (40 km per year for 5 years, as a suggestion, would double the effort made regarding the SONEES capital). Communication campaigns about water savings should be intensified. Communication media could be improved probably meaning the establishing of more resources. In the particular case of administrations, it will be essential to continue the action put underway to avoid the enormous waste observed so far. Note that this communication is a public responsibility. At the time of institutional reform, it would be worth looking at the organization in charge of this matter. If the Senegalese authorities want the future distribution company to handle this communication, for budget-related reasons, it must be provided for explicitly in the terms of reference of service delegation. o In another framework, it should be noted that a survey into the reuse of water for vegetable gardeners is currently underway. Implementation conditions: As far as water savings are concerned, it would appear that efforts should be made to maintain and service internal networks of large consumers to avoid waste. First, it is desirable to identify the urgent work needed, in conjunction with the Patrimony Directorate. Furthermore, it is advisable to reinforce the supervision and servicing of ALG1 manholes. In the crop production and market gardening area between Thies and point K, several manholes have been vandalized and cause leakage, as observed by reconnaissance in the field. The reuse of waste water is an interesting project, first to save drinking water and second, to prevent water from being reused in poor conditions as is currently the case in the semi-urban area of Dakar where the irrigation of vegetable gardens using untreated waste water is causing problems of hygiene. Another use of treated waste water is that of the watering of the Dakar parks and gardens, but only in a small proportion (approximately 20 m3/d or less than 1% of the Camberene effluent). A recent survey by BADJI/DELANOE/ADIN in October 1994 and work by JICA (1994) shows that the reuse of the waste water from the Camberene station could be suitable for irrigating certain crops, for watering trees (RN 1), or landscaped areas not used by the public. Other uses are possible (watering the Camberene golf course, irrigation of vegetable garden crops) but call for complementary treatment of the treatment station waste. If there is proof of the demand, the technical-economic conditions are not detailed enough and draft projects would need to be established on the scheduled sites: Camberene treatment station, installations at the Niaye station in Pikine. However, JICA sets the cost at 5800 MF CFA for an installation from Camberene serving the nearby crop-producing area (10,000 m3/d) and the Mbaw rewooding perimeter. 7.2.2. Monitoring of the quality of distributed water This is not a matter of the general inspection of the quality of the water distributed by SONEES but of two specific cases: the water drawn off the THIAROYE quaternary sand ground-water table and the problem of chlorine consumption by the NGNITH water. - The THIAROYE ground-water table nitrate contents are well above the permissible concentrations: values between 144 and 285 mgtl have been measured. This observation results from a problem of the chronic pollution of the ground-water by the infiltration of residual water from the PIKINE-THIAROYE built-up area. Although this has not actually been observed, it is possible that it could have non-conformning bacteriological results (the chlorine factors applied are fortunately relatively high). So far, SONEES was able to mix the THIAROYE ground-water with water from the Eocene limestones in a proportion of I volume of THIAROYE water to 2.5 volumes of unpolluted water thus diluting the nitrates to approximately 40-80 mg/l while the OMVS prefers a limit value of 50 mg/l. This trend towards the pollution of the THLAROYE ground-water by nitrates could become greater in the forthcoming years because the population is growing fast and the drainage infrastructure is not keeping pace with this increase. Even when drainage becomes a priority, in many districts it is impossible to install networks because of the way the streets are set out. These districts will have to be restructured beforehand. In order not to lose the THIAROYE resource during the interim phase, it would be worthwhile diluting it more. The design of the distribution restructuring work should therefore be an opportunity of thinking about this possibility. Eventually, this ground-water table will have to be abandoned as a drinking water supply. However, the drawing off must be maintained because the drop induced by the bore holes will probably prevent the flooding of certain districts built in the "niayes". - As far as the quality of the treated water at NGNITH is concerned, the problem involves the risk of unwanted substances derived from chlorine appearing (expected of being carcinogenic). The chlorine requirement of the water is effectively high, leading BETURE to propose a test and analysis campaign (contents in haloforms, depending upon the levels of chlorination). This survey is interesting and should be considered as an accompanying measure for the project. However, we should relativize the sanitary risk of chlorine derivatives resulting from the insufficient vaccination of the water. Implementation conditions: The water from the Thiaroye ground water, charged with nitrates, is currently being used to the extent of 7200 m3/d (Dec. 94) and is then diluted with water drawn from the ground water at Pout and Sebikotane in a proportion of one volume to 2.5, i.e. a dilution of the nitrates to 40-80 mg/l. This dilution is now more emphatic since the construction of a 600 mm diameter pipe leading from ALG1 (ND 600 HLM). As far as the campaign of analyses concerning haloform contents is concerned, if it is decided upon, it appears to be preferable to use the means of a private laboratory because SONEES does not have appropriate means of analyses and the acquisition of such equipment is not justified for the time being. REARkKS ABOUT THE SANITARY IMPACT OF THE PROJECT: The measures that might be proposed in this area are positive: they are not aimed at compensating an impact but rather on taking advantage of the project for sanitary improvement. The survey of the DAKAR drainage master plan is underway. However, it cannot be implemented within the 3rd water project, no matter how desirable, because of the financial stakes involved. Nevertheless, it will be seen that the greatest sanitary problems in the region result from the insufficiency of the drainage system at RUFISQUE. That is why it is particularly advisable to consider its financing during the 3rd water project. Located 20 km from Dakar in the bay of Hann, Rufisque is one of the oldest towns in Senegal and will be concerned by the interim phase of the third water project because it is served by the current ALG1 pipe. The population totals 118,000 inhabitants (1991 figure) and water consumption is 2,500,000 m3/year (SONEES 1992 figure). As mentioned above, interim phase impact will be positive regarding sanitary matters because of the better water service for the population. The negative impacts, due to the increase in dumped waste water volumes will affect the hydraulic area (the projections made refer to a volume of 160 I/d/inh in 2010 against 100 I/d/inh in 1993 - source: study into the drying out of DAKAR; JICA 1994) and to a lesser extent, regarding the individual polluting charge plan (from 47 g DBO5/inh/d to 60 g DBO5/inh/d in 2010 - source: JICA 1004, and the strategy plan). With respect to the Dakar built-up area, where there is an existing drainage project incorporating the water consumption hypotheses of the "Cayor Canal" phase, Rufisque is specific in that the drainage situation there is particularly degraded and is causing sanitary problems. Increased hydraulic and polluting charges in the interim phase will therefore have stronger impact there, justifying the implementing of a priority drainage project on this site. The drainage master plan drawn up in 1991 measures as initial priority the evacuation of waste water from old Rufisque (30 hectares) and the construction of a gravity network with evacuation of effluent to a purification zone, estimated as costing between 1.5 and 2 billion F CFA (before devaluation). 7.3 COMPENSATORYMEASURESAND RECOMMENDATIONS CONCERNING IMPACTS DUE TO THE PIPE LAYOUT The layouts of the main pipelines have been defined by GERSAR/BETURE in terms of the detailed preliminary project. They were the subject of end to end reconnaissance by SETUDE to estimate the impacts and describe the compensatory measures announced below. As mentioned in chapter 5, impact is relatively low on the neighborhood considering that the pipes are underground and that the ground equipment is almost all located in areas already devoted to the WS. 7.3.1. Safety of persons and access to trench General measure: the safety of persons during work on site requires that access to the trench be carefully protected: - on all areas near the villages (GUEOUL-THIES line); - on the urban parts of the THIES-POINT B line; - and in built-in areas where large pipes have to be replaced (distribution section). Implementation conditions: The contractor terms of reference must absolutely draw the attention of the contractors to this matter and indicate the minimum arrangements needed to protect the site, in particular when crossing dense urban zones between Thiaroye and Pikine where the pipes will be run causing hindrance to nearby residents (difficulties in getting around and access to homes, difficult pedestrian traffic). 7.3.2. Skirting around the villages on the GUEOUL-THIES line On the GUEOUL-THIES line, the villages mentioned in the section about impacts must be skirted around on the right. Note that this measure has already been taken into consideration by BETURE who are drawing up the detailed draft project. Preferably, work on site should not take place during the crop cultivation period. Implementation conditions: This measure is now taken into consideration by BETURE/GERSAR in its detailed draft project. 7.3.3 Rewooding in the Poa2tforest The running of the pipe through the Pofit forest will induce minor impact. Because this is a listed area, rewooding is nevertheless recommended to improve a site where high utilization pressure is applied. Implementation conditions: Refer to the impact study by BETURE/GERSAR for rewooding (400 plants/ha of Proposis juliflora) on a site to be defined, amounting to 8.5 ha and a cost of 3.7 MF CFA. Rewooding must be defined in conjunction with the Dakar woods and forests department in particular regarding the choice of species (monospecific crop?). 7.3.4. Reconditioning of the areas crossed within the Mbaw rewooding perimeter The MBAW rewooding perimeter, 771 ha in surface area, was created by the decree of January 1, 1950. This perimeter includes local species and species introduced by man (eucalyptus, cashew nut) and is now crossed by the BONNA and the layout of the new conduit will be parallel to it over a distance of 2.5 km and within a right of way deliberately reduced to 15 m. Implementation conditions: To compensate for the impact of the site in crossing the perimeter, it is proposed to repair the location after work within the limits of the constraints caused by the creation of a working road which does not exist for the time being. If it is decided to construct this road, which will not allow the site to be repaired, it will be necessary to think in terms of increasing the perimeter over a surface of 3.7 ha (right of way 15 m over 2.5 kim) and to rewood it. After consulting with the Dakar Water and Forests department, the garden-raised cashew nut tree species appear to be that best suited for the repair of the site or for rewooding. The cost of the measure is calculated on the basis of rewooding a 3.75 ha perimeter for a total of 1.5 to 1.9 MF CFA i.e. 400,000 to 500,000 F CFA/ha. In addition, arrangements must be made to water the young plants during the dry season. 7.3.5 Compensation budgetfor narket gardening operating losses on the THIES- POINT B line For the THIES-POINT B line, it will be necessary to provide a compensation budget for vegetable gardening operational losses. If possible, the period of work on site, must be chosen to minimize the impact. Finally, it will be essential to avoid felling baobab trees except for 3 or 4 of them which cannot be avoided. Implementation conditions: Refer to the BETURE/GERSAR impact study for compensation calculated on the basis of a cost of 465,000 to 1,060,000 F CFA/ha during the first year and 60% of that cost during the second year. 7.3.6. Integration of the Mamelles reservoir in the site As far as the MAMELLES reservoir is concerned, the BETURE survey has already pointed out that the design must be integrated into the site (cf 5.2 and 62). Because of the large size of the installation, we think that this concem (a 25,000 m3 reservoir during the first stages of the interim phase) has to be complied with. Implementation conditions: The site at which the Mamelles reservoir is to be located is one of the highest sites in the town where there is already a storage installation, under satisfactory conditions regarding landscaping integration. The new reservoir, rectangular and L-shaped in its design, will be integrated into the site which could be characterized by its natural gentle slope from elevation 75 NGS to elevation 50, opening onto the Ngor road. Accordingly, backfill must be used to the utmost in order to smooth the facades that are visible from the Ngor road in particular (height 6 m including the vegetal cover to be provided over 75 m which is an essential complement for the integration of the construction into its environment). 7.3.7. Crossing the THL4ROYE-PIKINE dense urban area The characteristics of this urban area appear as follows: - there is no control on the dwelling situation and there is structure to the real estate situation (according to the information gathered from the Urbanism department, ten or so private and large deeds are concerned by the layout in the first zone at Pikine while these areas are entirely occupied by irregular tenants). The remainder of the land is multinational (non-registered land). - the average surface of the irregular lot is 150 i - this zone is in a state of perpetual change with new irregular constructions appearing rapidly, - the zone has been the subject of many restructuring projects such as the passage of a motorway, the creation of new roads (GTZ project) and the Medina Fass restructuring project. The pipe layout, defined in the detailed draft projects on maps to 1:1000, runs through the Thiaroye-Pikine urban area at the exit of the Mbaw perimeter in the following sequence: initial dense irregular urban zone at Pikine for approximately 800 m, unbuilt parts of the Faidherbe camp, second dense urban zone at Guinaw rails then following of the existing ALG layout after the SNCS tracks. Over this route, which was the subject of detailed reconnaissance during this assignment, only two regular dwellings need to be destroyed in the last part mentioned above (a home and a garage located within the right of way of the existing ALG), as provided for in the detailed draft project. Conversely, the construction of the site will itself be a source of extensive nuisance for the populations concerned because of the restricted possibilities of travel and access to houses along the layout. This will probably cause specific technical arrangements to be made in terms of the contractor consultation book (sheet piled excavations, fast covering of the trenching, maintained access to the homes, speedy work). Implementation conditions: Insofar as the trenching techniques allow a layout without demolition in the uncontrolled area, the steps to be taken to limit impact upon the people concern the minimizing of nuisances i.e. the speed of execution of the site and the safety of the inhabitants. As pointed out above, the specific technical provisions required of the contractor will play a determiing role. For the homes affected by the running of the conduit through the regular area, after the SNCS track, a compensation budget will be needed, estimated at 10 MF CFA. APPENDIX 1 Memos drafted at DAKAR subsequent to the mission of June 1994 APPENDIX 2 Memos drafted at DAKAR subsequent to the mission of February 1995 APPENDIX 3 Synthetic evaluation of ground-water tables 3.1. Infrabasaltic sandy ground-water table 3.2. THIAROYE quaternary sandy ground-water table 3.3. SEBIKOTANE Palaeocene limestone ground-water table 3.4. POUT Palaeocene limestone ground-water table 3.5. Maestrichtian sandy ground-water table 3.6. North Coastal ground-water table APPENDIX 3.1. - INFRABASALTIC SANDY GROUND-WATER TABLE GENERAL DESCRIPTION This concerns quatemary marine sand, overlaid by basalt flows. The ground-water surface is free, fed either directly by fractures due to basalt shrinkage or from the East, where the basalt layer disappears in places. Over-exploited at certain times with respect to the refilling rate (years 50 and 70-85), the salty advance has progressed in the West and South-East. It appears to have become stabilized for ten years or so. DRAWING-OFF Drawing-off has been relatively stable since 1985 at around 17,000 to 19,000 m3/day. Drawing-off was by SONEES in 8 bore holes operating at 80 to 170 m3/h. PIEZOMETRY 18 piezometers were followed up in 1991. Following a rise between 1989 and 1991, the ground-water has dropped over the last few years apparently because of the rainfall which is again deficient. The entire central part of the aquifer, and the part close to the North Coast are at an elevation of less than 0. SYNTHESIS OF REMARKS AND OBSERVATIONS MADE IN THE DIFFERENT STUDIES The ground-water supply appears to be closely connected to rainfall. The reduction in flows sampled at 13,820 m3/d had been recommended by BETURE in 1986. Thanks to satisfactory weather conditions, the flow-rates were maintained at 18,000 m3/day on average. The flow-rates set for the interim phase presuppose that satisfactory rainfall should be maintained. FORECASTS Exploitation at 20,000 m3/day during the interim phase. Installation of three new piezometers. * Eventually, ("Cayor Canal" phase), the flow will be reduced to 14,200 m3/day and will only be drawn off at peak periods. RECOMMENDATIONS To monitor very attentively the chemical evolution of this compartment. To cut down on flows if the rainfall becomes deficit. APPENDIX 3.2-QUATERNARY SANDY GROUND-WATER TABLE GENERAL DESCRIPTION The quatemary sands on the Cap Vert peninsula support a free table fed directly by the impluvium. It extends from 2 DAGOUDANE-PIKINE to KAYAR over a surface area of around 400 kmI. We can distinguish between the two areas where it is utilized: the Beer Thialane basin and the Thiaroye basin. DRAWING-OFF The THLAROYE basin is used for DWS. Drawing-off is slightly on the decrease, dropping from 10,000 m3/day before 1985 to 7,800 m3/day in 1992. SONEES has 7 wells. The Beer Thialane basin is reserved for agricultural purposes and is operated at a rate of 8000 m3/year on average. PIEZOMETRY 48 piezometers were monitored in 1991. There has been a relative stability in the levels since 1985 with even a rise in the south and west. SYNTHESIS OF REMARKS AND OBSERVATIONS MADE IN THE DIFFERENT STUDIES The Thiaroye ground-water table, of interest to the DWS is not exploited to its maximum potential which is estimated at 10,500 m3/year. Beyond that, there would be a risk of salty invasion from the South. However, the most important remark concems the water pollution due to nitrates with contents ranging from 150 to 300 mg/l (value recommended by the WHO: 50 mg/l). This phenomenon is probably due to urban development and to the lack of drainage in many districts of PIKINE- THLAROYE. To be consumable, the water has to be mixed with water containing few nitrates. FORECASTS The SONEES drawing-off levels are maintained at 8300 m3/day. A piezometer will be installed South of Thiaroye at DLAMAGUENE in order to control the progression of the salt water influx. In the Cayor Canal phase, 7500 m3/day but only in the peak periods. RECOMMENDATIONS It is essential to closely monitor nitrate contents and the bacteriology and to inform the population about the risks involving infants and expectant mothers. During the CAYOR CANAL phase it would be preferable to abandon the resource entirely as a source of drinking water. Drawing-off should be maintained in the ground-water in order to maintain the "niayes" that have been urbanized above water. APPENDIX 3.3. - SEBIKOTANE PALAEOCENE LIMESTONE GROUND-WATER TABLE GENERAL DESCRIPTION These are more or less karstified limestones forming a compartment located between two major North-South faults, extending over 25 kIn from the ocean in the North to the ocean in the South. Located between a roof of marnes, the ground-water table is captive and fed from the East by exchanges with other aquifer systems and in the extreme South (where it becomes free and also receives rainfall input). It was highly exploited in the beginning of the 60's (35,000 m3/day) and the southern part has been invaded by salt water. DRAWING-OFF Drawing-off for DWS is stable at 25,000/27,000 m3/day since 1985. SONEES has four bore holes with flow-rates ranging from 200 to 400 m3/h. Agricultural drawing-off amounts to approximately 4000 m3/day. PIEZOMETRY AND SALINITY 9 piezometers have been installed. The ground-water table is dropping, on average, 50 to 60 cm per annum. The PANTHIOR dam in the South is used for occasionally resupplying the ground-water after the winter period and forming a temporary piezometric dome which probably slows down the salt water invasion. Nevertheless, a bore hole was stopped (F4) and the chlorine contents measured on some of the piezometers are worrying (change from 35 mg/l in 1987 to 291 mg/l in 1993). SYNTHESIS OF REMARKS MADE IN THE DIFFERENT STUDIES The ground-water table was modelled by ARLAB in 1986. The model gives piezometry results that are far more pessimistic than real conditions and it is difficult to take them into consideration. Nevertheless, the ground-water is obviously over-exploited and the progression of dry residue is alarming. 3 BETURE recommended in fact that the water drawn off should be reduced to 17,500 m /day as of 1990. If necessary, we will have to replace the bore holes in this ground-water table by bore holes in the Maestrichtian tables. In a study of March 1983, BRGM underscores the role played by the PANTTHOR dam in resupplying the table and in preventing the advancing of the salt water in the South. On the other hand, BETURE considers that its efficiency is barely satisfactory (mission 0, appendix 1). FORECASTS Maintaining drawing-off at 25,000 m3/day Installing a new piezometer In the Cayor Canal phase, decreasing to 17,500 m3/day RECOMMENDATIONS The resources truly "provisional" at this level of operation thus justifying, among other motivations, acceleration of the implementation of long-term DWS projects * In the event of there being any doubts, the refilling role of the PANTHIOR dam must be maintained, by flushing it every two years. APPENDIX 3.4. - POUT PALAEOCENE LIMESTONE GROUND-WATER TABLE GENERAL DESCRIPTION These are low karstic limestones forming a narrow but long compartment 50 km long from North to South (from the ocean in the North to the ocean in the South), confined within faults on East and West. The table is free south of FOULOME and captive in the North (covered by marnes). In the North, it is fed by underground supplies from a fossil valley (KEURMAYE). In the South it is fed by infiltration (low or zero under the current climatic conditions). The southern part was over-exploited in the 60's (up to 17,000 m3/day) inducing a risk of saline invasion. The northern part has only been exploited since 1978 but in an increasing manner. EXPLOITATION SONEES has two bore holes operating at 100 m3/h at South Pout and 4 at 160 to 300 m3/h in North Pout. There is a slight decrease in the amount drawn off since 1983 from 27,000 m3 in 1983 to 24,000 m3 in 1991 and 17,000 m3 in 1992 because of the commissioning of new bore holes in the underlying Maestrichtian surfaces. The ground-water table is also solicited for agricultural water. BETURE mentions 5000 m3/day in its 1986 report (volume 1). PIEZOMETRY AND SALINITY The ground-water is monitored by 16 piezometers. In the North it is dropping 0.6 m per annum. Salinity is quite high near the TAMNA lake but the water is otherwise of good quality. In the South, drops are small and irregular with even rises in places. Conversely, in the southernmost part, the chlorine content has increased from 120 to 275 mg/l between 1985 and 1993. SYNTHESIS OF REMARKS AND OBSERVATIONS MADE IN THE DIFFERENT STUDIES The ground-water table was modelled by ARLAB in 1983 and 1986 (particularly complete multi-layer model with 3000 grid links, enormous work on the breakdown of the hydrogeological parameters). The possible drawing-off amounts depend upon the drawing-off from the Maestrichtian layers. A model scenario with the following SONEES drawing-off has been made: 20,000 m3/day up to 1992, 11,000 m3/day beyond, providing for a loss of 5 m by the horizon 2010 and with no significant encroachment of the salt water. FORECASTS In the interim phase, maintained operation at 20,000 m3/day Installation of three additional piezometers In the Cayor Canal phase, reduction to 11,000 m3/day RECOMMENDATIONS * Because operation is being maintained for several years at a level higher than forecasts, we must stand by to adjust the amounts drawn off (to the benefit of the Maestrichtian layers) if piezometric supervision discloses excessive drops in the North or if the salty encroachment continues in the South. * The operational level of this compartment is necessarily provisional, suggesting that particularly fast implementation of subsequent phases of the DWS are to be brought to the forefront. APPENDIX 3.5. - POUT MAESTRICHTIAN GROUND-WATER TABLE GENERAL DESCRIPTION This immense deep table is powerful (500 m thick) and captive. It extends over a large part of Senegal up to the Senegal river. It receives water coming from the Palaeocene limestone ground-water tables with a number of inlets at the limits. It was only recently put into operation (1971) and has slowly but regularly built up to 1991. This table has made it possible to increase the DAKAR resources for the emergency phase. DRAWING-OFF SONEES had 8 bore holes before 1986. Ten new ones have just been put into service as part of the urgent phase (1992). The flow-rates are included between 125 and 200 m3/h. The samples thus increased from 37,000 m3 in 1991 to 62,000 m3/day in 1992. Other drawing-off is considerable, estimated in 1985 at 50,000 m3/day (TAIGA, THIES, M'BOUR, etc.). PIEZOMETRY AND SALINITY 14 piezometers were in operation in 1991. Evolutions vary enormously from a I m per annum drop to a slight rise. The piezometric readings in May 1994, supplied by SONEES, indicate that the drop is far more emphatic since exploitation was increased: almost everywhere the drop is approximately 1 m per annum and even attains 2 m per annum on one piezometer. The quality of the water is very good except in the South where salinity is increasing because of resupply by the Palaeocene layer itself affected by the invasion of salt water. SYNTHESIS OF REMARKS AND OBSERVATIONS MADE IN THE DIFFERENT STUDIES The ground-water table was modelled by ARLAB in 1986. Piezometric forecasts are, however, too pessimistic with respect to reality (expected drop of 3 m per annum with current exploitation levels). Therefore it cannot be used as a reference. BETURE recommends that the drawing-off be reinforced if other resources drop at DAKAR. FORECASTS In the interim phase, drawing-off is increasing to 73,000 m3/day 6 additional piezometers will be installed In the CAYOR CANAL phase, drawing-off can be maintained RECOMMENDATIONS A piezometer must be added in the southernmost section to replace SE3 1 which appears to be deficient. It is absolutely essential to protect this aquifer from the invasion of salt water in the South and to do this, stop the drilling into the southern POUT limestones if the trend towards salinization continues. COMPLEMENTARY REMARK CONCERNING THE THIES MAESTRICHTLN GROUND-WATER TABLE The compartment comes in for less extensive exploitation for the city of THIES: 8 SONEES bore holes for scheduled production of 13,700 m3/day in 1993. The ground-water level is dropping 80 cm per annum. The interim phase provides for exploitation of two additional bore holes bringing production capacity to 17,000 m3/day. BETURE underscores that the hydrodynamic qualities of this part of the aquifer are poor and suggests a rapid drop in subsequent drawing-off. Because there is a definite lack of data gathered as part of this mission, no conclusions or recommendations can be made at this stage. APPENDIX 3.6. - NORTH COASTAL GROUND-WATER TABLE GENERAL DESCRIPTION The North Coastal ground-water table extends from TIVAOURE to SAINT-LOUIS. It consists of two lithological sets, quaternary sands between the RN2 highway and the sea and Eocene limestones east of the highway. The ground-water has a free surface which is only supplied by rainfall and only when that rainfall produces run-off (North part of Senegal where the average rainfall is around 300 mm/year). There are two major hydrological characteristics to be borne in mind: the existence of a piezometric dome along the South Coast, protecting the southern part against salty invasions and, to the contrary, the definite risk of salty invasion in the North towards LOUGA. The water quality is good. DRAWING-OFF Operation is by SONEES for the CAP VERT DWS since 1978. The exploitation level increased from 14,300 m3/day in 1978 to 27,000 m3/day in 1992 (4 bore holes at KELLE, 1 bore hole at KEBEMER, connected to the ALG pipe). To this we must add approximately 6000 m3/day drawn off by SONEES for LOUGA, BAMBEY, KHOMBOLE, DIOURBEL. It is also used to supply the ICSs, the Senegal Chemical Industries (irregular volumes of between 4000 and 9000 m3/year). Agricultural supplies are estimated as consuming 2400 m3/day. Drawing-off from the bore holes supervised by DEM are around 3000 m3/day. Finally, village use excluding the DEM and SONEES bore holes represent a roughly estimated total of 25,000 m3/day. In all, drawing-off comes to approximately 70,000 m3/day. PIEZOMETRY 49 piezometers were monitored from 1986 to 1991. They indicate a drop in the piezometric dome of 50 cm/year since 1975 and a drop in the ground-water near the KELLE bore holes of 30 cm per annum with, elsewhere, a drop along RN2 amounting to 10 cm/year. SYNTHESIS OF REMARKS AND OBSERVATIONS MADE IN THE DIFFERENT STUDIES This essential paragraph is developed in the main text of the study (basically it concerns the BRGM 1992 study). PERSPECTIVES * Increased drawing-off by SONEES amounting to 34,800 m3/day on average from the 11 bore holes at 125 to 150 m3/hour. * Installation of 6 piezometers especially in the North-Western part that is vulnerable to a salty invasion. * In the Cayor Canal phase, 3 different exploitation scenarios can be considered: maintained operation at 34,800 m3/day, reduction to 25,200 m /day or shutdown of over- exploitation. RECOMMENDATIONS This essential paragraph is developed in the main text of the survey. The recommendations concem: * the need to define a frame of reference for the piezometric monitoring of the ground-water table; * the need to choose bore hole locations as far as possible from the villages; * the need to provide for the deepening of the wells or bore holes in the villages which will dry up because of the start-up of the new bore holes. APPENDIX 4 Maps of neighboring effects 97 APPENDIX 5 Reproduced documents 98 Appendix 5.1 Synthesis of the texts concerning the area of water (ERNSTand YOUNG) 99 Appendix 5.2 Extracts of the Northern coast water resources re-evaluation survey conclusions (May 1992) 100 Appendix 5.3 Example of six-monthly report on piezometric follow-up (SONEES - laboratory) 101 Appendix 5.4 Renewal of networks and distribution of leakage (SONEES - Operation Management) 102 Appendix 5.5 The water sales policy at the fountain points (SOAMES - Operation Management) 103 Appendix 5.6 Social connections policy 0 97