About UNEP
United Nations Environment Programme
Division of Technology, Industry and Economics
top image
space space space

Newsletter and Technical Publications

<Sourcebook of Alternative Technologies for Freshwater Augumentation
in Small Island Developing States>


Annex 2

Design of Infiltration Galleries


Some of the main aspects to be considered in the design of infiltration galleries are outlined in this Annex.

Figure 46
Figure 46. Freshwater lens and galleries on Bonriki Isaland, Tarawa Atoll, Republic of Kiribati.
(A) shows details of galleries and monitoring boreholes, (b) shows a cross section of the lens based on salinity measurements in 1980.

Some additional details have also been added based on recent experiences with galleries, particularly in the Cocos (Keeling) Islands, Indian Ocean.

Figure 47

Figure 47. Freshwater lens and galleries on Home Island, South Keeling atoll in the Cocos (Keeling) Island: (A) Overall plan of island, (B) Details in area of housing an dinfiltration galleries.

Technical Description

The exact layout of infiltration galleries is dependent on the boundary of the freshwater lens and existing land uses, especially the location of permanent structures. Hence, layouts should tend to be parallel to edges of lenses and not perpendicular to them, except near the centre. Linear systems are probably the best approach on islands with limited lateral extent, and where the freshwater zone is relatively small. On wider islands, radial or cruciform patterns may be appropriate. Examples of layouts of galleries are shown in Figures 43 and 44, respectively, for Bonriki Island on Tarawa Atoll, Kiribati, and Home Island on South Keeling Atoll, Cocos (Keeling) Islands, Australia. Where possible, areas with certain land uses should be avoided. These include heavily vegetated areas; agricultural areas, especially where agricultural chemicals are used or where pits are dug to the water table (e.g., taro pits); and, areas of residential and industrial development. In some cases, infiltration galleries must be fitted to existing land use and other measures taken. On Home Island (Figure 47), the galleries had to be laid between existing houses which had earlier been built over the only freshwater lens on the island. The galleries on Home Island replaced earlier dug wells and simple extraction systems consisting of pumped wells with very short lateral pipes extending from their bases. A chlorination system was necessary to ensure the water was safe for potable purposes. In contrast, open areas are ideal for construction of galleries and include airfields, and sports fields and recreational reserves.

Galleries should consist of some form of horizontal, permeable conduit to allow water to infiltrate from the surrounding saturated zone. Suitable conduits have been constructed using slotted PVC pipe; porous concrete pipes made from no-fines aggregate; and, non-jointed hollow concrete blocks. Conduit systems considered to be generally unsuitable are steel pipe systems, because of possible long-term corrosion, and open trenches, because they are susceptible to surface pollution.

Typically, infiltration galleries consist of slotted PVC pipe conduit systems laid below the water table at or close to mean sea level. These have the advantage of being easily cleaned if required. If PVC pipes are chosen, pipes should be of 100 mm or larger nominal diameter to ensure that they can be cleaned out if necessary in the future; joints should be either of the rubber-ring or the solvent-cement type (however, as permanent jointing is not required, pipes could also just be pushed firmly together to prevent the entry of sediments between sockets and spigots); pipes should be at least Class 6 water-supply grade for structural reasons, slotted, and laid horizontally, in straight line sections to assist with cleaning; synthetic fabric around the slotted pipes should not be necessary and may in fact cause clogging of slots; and, caps should be fitted at the ends of gallery pipe sections to prevent the entry of sediments. If available, selected "no-fines" gravel should be backfilled around the pipes to prevent the entry of fine sediments into the infiltration pipes. However, as such gravel is often unavailable, it may be necessary to backfill with the excavated material and "develop" the gallery after construction. The "development" procedure consists of pumping the gallery at a higher rate than under normal operating conditions with the aim of drawing fine material into the pipe, from which it can then be cleaned out. This should leave a filter zone around the pipe consisting of particles which are too large to be drawn into the pipe. Also, an impermeable barrier such as thick polythene sheeting should be laid in the trench at a height of about 100 mm to 200 mm above the top of the pipe to assist in minimizing further downward movement of fine particles into the zone around the pipe (Figures 46 and 48).

Figure 48

Figure 48. Detailed cross section through an infiltration gallery constructed from horizontal slotted PVC pipes. Vertical concrete cylinders have been used for the pump well and access manholes. This design was used for six galleries on Home Island, Cocos (Keeling) Islands, Australia.

Both manual and mechanical excavation methods can be utilised for gallery construction. The choice of method, which may include a combination of the two, is dictated by local factors including the degree of difficulty of the excavation and the cost and availability of manpower and equipment. Some important factors to be considered during construction include, if possible, excavation in short stages to minimise the amount of open trench at any given time, which reduces the amount of dewatering required, and limits the amount of erosion of disturbed soils and possible contamination of the water source. Care must be exercised in the use of dewatering pumps as over-pumping can lead to local intrusion of saline water; pumping rates should be just sufficient to enable excavation below the water table, and subsequent laying of conduits and backfilling operations. Also, salinity tests should be performed regularly as excavation advances to ensure that significant salinity increases are not detected either due to encroachment into higher salinity groundwater zones or to dewatering at inappropriately high rates. Tests should be conducted at about 20 m intervals during excavation. If high-salinity areas are detected, the gallery alignment or overall length may require adjustment.

Conduits should be laid at such a level both to prevent dewatering during low tides and saltwater intrusion during high tides or periods of peak freshwater use. These problems may result if the conduits are too high or too low. It is suggested that the inverts of conduits should be set at between 100 mm and 300 mm below mean sea level (MSL), depending on the head of water above mean sea level (Figure 48). No allowance for possible long-term sea level changes is suggested because of uncertainty of the magnitude of changes. If sea level does rise, as is most likely within the next 50 years, infiltration gallery conduits would be slightly lower within the freshwater lenses than at present. This should not necessarily be a problem, particularly if there is an accompanying increase in precipitation which could increase the thickness of the freshwater lens. A more substantial effect on some small low-lying islands is the possible long-term inundation of certain areas. However, if the inundated areas are away from the freshwater lenses, the amount of disturbance could be minimal.

Based on Darcy's Law for groundwater flow, an approximate solution method was developed for estimating the horizontal length of gallery systems for the Cocos (Keeling) Islands (Falkland, 1988). The length of a gallery, Lg (in metres), can be expressed as:

Lg = W.Qg / 8.k.D.h (1)

W = width of extraction area (m);
Qg= pump flow from gallery (m3/day);
k = permeability (m/day);
D = approximate mean thickness of freshwater zone (m); and,
h = allowable drawdown (m).

The extraction area for a particular gallery is assumed to be that area of the lens influenced by the gallery. Based on the areal extent, A (with length, L, and width, W), of an assumed extraction area, the length of a gallery, Lg, can be equated to L - W. This implies that gallery is shorter than the length of the extraction area, L, by half the width (W/2) at each end. Within this area, the extraction or pump flow from the gallery, as defined above, should be proportional to the sustainable yield of the lens divided by the lens area. This can be expressed as:

Qg = A.QS / AL = (Lg + W).W. Qs / AL (2)


A = extraction area (m2) = W.L
QS = sustainable yield of entire freshwater lens (m3/day or kL/day), and
AL = area of lens (ha).

Methods to assess the sustainable yield of freshwater lenses are outlined in Annex 4. The length of gallery, Lg, can thus be obtained by simultaneous solution of equations (a) and (b) for a given set of parameters.

Pipe slots should allow the passage of water but should minimise the passage of sediment particles. It is recommended that the slot width not exceed 1.5 mm. Entrance velocities at the slots should be below a certain maximum value to prevent clogging. Huisman (1972) mentions that, for the design of radial collector wells, which are similar in design to horizontal gallery pipes, the maximum entrance velocity (v) may be calculated using the square root of the permeability (k) of the surrounding aquifer material divided by 60 (both v and k are expressed in m/sec). This design criteria is considered too conservative in most cases. Rather a velocity of double this value, which is often used for permanent vertical well screens, is suggested. An example of a slot pattern for 225 mm PVC pipes used for the Vaipeka gallery extension on Aitutaki is shown in Figure 49, and Figures 46 and 47 provide an outline of design options that have been found useful in galleries on Tarawa and Home Islands.

It is also good practice to install manholes (or access holes) at spacings not exceeding 100 m to enable conduits to be desilted if necessary. Manholes should be about 1 m in diameter and be constructed using suitable, corrosion-resistant materials such as prefabricated cylinders made from fibreglass, concrete, ferrocement, or, where available, local materials such as coral slabs. Joints between pipes and manholes, and between manholes and manhole bases, should be well sealed, preferably with concrete, to prevent water from entering the system; and, manholes should be fitted with sealable lids. The top of the manholes can be covered if required. Further, in addition to the manholes which are primarily installed for inspection and maintenance purposes, it is useful to install testing points at selected locations by inserting PVC tees into the gallery pipes, and extending the tee to the surface. To allow the entry of a salinity probe, the pipes should be about 50 mm in diameter.

Generally, one pump per gallery is sufficient. Pump wells or sumps can be constructed using the same materials as for manholes. Joints between the pump wells and incoming infiltration pipes and the pump well base should be properly sealed, as for manholes. These measures are important to ensure that the only water entering the system is through the slots in the pipes, and, hence, to prevent local overpumping and consequent excessive drawdown, and contamination from surface sources. Pump wells should be properly covered to exclude animals, birds and sunlight, and the rims of pump wells should be set about 0.5 m to 1.0 m above ground level to prevent local inflow during heavy rainfall to minimise the chances of contamination. The base (inlet) of pump suction well should be set so that it is at least slightly above mean sea level, although, in some cases, it may be necessary to set the inlets slightly below mean sea level to prevent the pump running dry at very low tides. Float switches should be set to turn the pump off when the water level approach the base of the well to prevent the pumps from sucking air

The types and sizes of pumps are dependent on the required yield, the head, and the length of the gallery. There is a trade off between pump rate per gallery and the number of galleries. The product of these two quantities must be no greater than the sustainable yield. Hence, it is good practice to set the pump rates at a rate somewhat lower than the calculated rate until the effects of long-term pumping have been monitored. Also, to minimise disturbances to the freshwater lens, pumping should be continuous rather than intermittent, although this may not be possible if solar pumps (with no battery back-up) are utilised. In this case, pumping should be continuous throughout the pumping period. The prime consideration is to minimise the drawdown in the well due to the pump to prevent seawater intrusion. Mather (1975) suggests a maximum drawdown due to a pump of 30 mm (0.03 m) except for very thick freshwater lenses. Drawdowns of between 30 mm and 50 mm may be considered reasonable for most lenses and were used in the design of galleries in the Cocos (Keeling) Islands (Falkland, 1988). Similarly, there have been a number of approaches to the selection of pump rates for galleries.

Extent of Use

For freshwater lenses in the Trust Territory of the Pacific Islands (now Federated States of Micronesia), Mink (1986) suggested a pump rate of 0.10 gal/min/ft (or 1.8 m3/day/m) of gallery. Examples of pumping rates from operational galleries on coral atolls include:

  • A total pumping rate from seven Kwajalein galleries with a combined length of 2 120 m of about 340 m3/day in 1980 (Hunt and Peterson, 1980; Peterson, 1984) -- This pumping rate is equivalent to 48 m3/day per gallery or 0.16 m3/day/m of gallery. Using the sustainable yield estimate of about 520 m3/day (Peterson and Hunt, 1981), the upper pumping rate for the galleries would be approximately 0.25 m3/day/m of gallery.
  • A total pumping rate in 1992 from 23 Tarawa galleries at Bonriki (17 galleries: Figure 46) and Buota (6 galleries) with a combined length of about 6 900 m of about 1 300 m3/day, which is slightly less than the estimated sustainable yield (Falkland, 1992) -- This equates to an average pump rate of 56 m3/day per gallery or 0.19 m3/day/m of gallery. On an areal basis the pumping rate at Bonriki is equivalent to about 7.3 m3/day/ha, based on the total area of the lens.
  • A total pumping rate from six galleries on Home Island, Cocos (Keeling) Islands (Figure 47) of approximately 110 m3/day which is equivalent to the sustainable yield (Falkland, 1992) -- This pumping rate equates to 19 m3/day per gallery or 0.06 m3/day/m of gallery.

From these data, it can be seen that the average pumping rates for the galleries on Kwajalein and Tarawa are approximately the same, while that for Home Island is about one third of that rate. It should be noted that the freshwater lens on Home Island is much thinner than at the other two locations. Examples of pump rates from galleries on islands other than atolls include one from the Pacific and one from the Caribbean:

  • A total pumping rate from the Vaipeka gallery on the island of Aitutaki, Cook Islands (Figure 49), of about 480 m3/day or about 3.4 m3/day/m of gallery -- This is much higher than the rates for atolls as there is a sizeable catchment area for the gallery. The original gallery was constructed parallel with the shoreline at the base of a volcanic rise, 140 m in length, and a total pump rate, shared between two pumps, of about 5.5 l/s. The extension to the gallery is 130 m and will initially have one pump rated at about 3 l/s (Falkland, 1994, 1995). A second pump of the same capacity may be installed on the extension if monitoring proves this to be feasible.
  • A total pumping rate from the open trench galleries on the limestone islands of New Providence and Andros, Bahamas, of 120 m3/day per 300 m gallery (Lloyd et al., 1993) -- This is equivalent to 0.4 m3/day/m of gallery. However, recent modelling suggests that this pump rate is too high and that an upper pumping limit of 58 m3/day per 300 m of gallery (or 0.19 m3/day/m of gallery) should apply (Lloyd et al., 1993). This lower value is very similar to the rate adopted for Tarawa.

From the above discussion, gallery-pump rates for atolls and small limestone islands should probably be less than 0.25 m3/day/m of gallery, and, in the case of very fragile lenses, should probably be less than about 0.10 m3/day/m of gallery.

Information Sources


Tony Falkland, ACTEW Corporation, Australia


Falkland, A.C. 1988. Cocos (Keeling) Island Water Resources and Management Study. ACT Electricity and Water Report No. HWR88/12, ACT Electricity and Water, Canberra. 211 pp + appendices.

Falkland, A.C. 1992. Review of Tarawa Freshwater Lenses, Republic of Kiribati. ACT Electricity and Water Report No. HWR92/681, ACT Electricity and Water, Canberra.

Falkland, A.C. 1994. Review of Vaipeka Gallery Extension and Water Supply Needs. Australian International Development Assistance Agency, Canberra

Falkland, A.C. 1995. Vaipeka Water Gallery Extension Project, Aitutaki, Cook Islands. Australian Agency for International Development, Canberra. 62 pp + appendices.

Huisman, L. 1972. Groundwater Recovery. Macmillan, London.

Hunt, C.D. and F.L. Peterson 1980. Groundwater Resources of Kwajalein Island, Marshall Islands. Water Resources Research Centre Technical Report No. 126, University of Hawaii. 91 pp.

Lloyd, J.W., Yu Yong Hong, and D.W. Peach 1993. The Management of Groundwater Abstraction in Islands Using Trenches: A Bahamian Example. In: Study and Modelling of Saltwater Intrusion into Aquifers, Proceedings of the 12th Saltwater Intrusion Meeting, Barcelona. pp. 489-503.

Mather, J.D. 1975. Development of the Groundwater Resources of Small Limestone Islands. Quarterly Journal of Engineering Geology, 8:141-150.

Peterson, F.L. and C.D. Hunt 1981. Groundwater Resources of Kwajalein Island, Marshall Islands. Water Resources Research Centre Technical Memorandum Report No. 63, University of Hawaii. 38 pp.

{short description of image}

Table of ContentsTable of Contents Next

  • Brochure
  • IETC Brochure

  • International Year of Forests
  • International Year of Forests

  • World Environment Day
  • ??????

  • UNEP Campaign
  • UNite to Combat Climate Change