Hard-rock aquifers in tropical regions: using science to inform development and management policy Stephen Foster Abstract An overview of progress during the past 30 years in the hydrogeologic understanding of groundwater in hard-rock aquifers of tropical regions is presented. Geographically, the paper concentrates upon and contrasts Tropical Africa and Peninsular India, where very extensive areas of weathered hard-rock aquifers occur, but its conclusions are more widely applicable. This scientific understanding forms the basis for a critical discussion of key policy issues for the development and management of the water resources of these aquifers, given their major importance for economical and sustainable water-supply provision, in the context of efforts to achieve the UN Millennium Development Goals for rural drinking water and improved livelihoods. Keywords Tropical Africa . Peninsular India . Groundwater management . Hard-rock aquifers . Groundwater development

Historical perspective on groundwater interests Looking back to the UN Drinking Water and Sanitation Decade The United Nations (UN) Drinking Water and Sanitation Decade (1981–1990) led to increased attention on the search for low-cost rural drinking-water sources, especially in the hard-rock terrains of Peninsular India and Sub-Saharan Tropical Africa (Fig. 1). Efforts were centred around three inter-related aspects: – Provision of low-cost, ultra-portable, lightweight percussion and pneumatic drilling equipment. – Technical innovation to reduce the cost and increase the reliability of waterwell handpumps. Received: 28 June 2011 / Accepted: 31 December 2011 Published online: 15 February 2012 * Springer-Verlag 2012 S. Foster ()) International Association of Hydrogeologists (IAH), c/o PO Box 4130, Goring-on-Thames, Reading, RG8 6BJ, UK e-mail: [email protected] Hydrogeology Journal (2012) 20: 659–672

– Initiation of hydrogeological research and surveys of weathered hard-rock aquifers, which occupy very large land areas in these (and other) regions. Leadership on the former two was mainly by multilateral donors—notably the United Nations Children’s Fund (UNICEF), the World Health Organization (WHO) and World Bank—however, various bilateral donors put considerable effort into improving the scientific foundation. Traditionally, dugwells had been used by rural communities to obtain water-supply for domestic use and livestock rearing in weathered hard-rock terrains, but they were highly prone to drying-up in the dry season or during extended drought, hence the interest in drilling waterwells deeper into the weathering profile than was technically and/or economically feasible with dugwells. The first hydrogeological research on the weathered crystalline basement (British funded in Malawi and Zimbabwe) in the 1980s led to the elaboration of a generic hydrogeological conceptual model (Chilton and SmithCarington 1984; Fig. 2), comprising the widespread occurrence at shallow depth of a fractured ‘saprock horizon’ capable of yielding groundwater to borewells or tubewells, overlain by a much less permeable ‘saprolite’ which was widely preserved and provided limited but useful groundwater storage, at least in areas of significant seasonal rainfall (> 750 mm/a). This had important implications for groundwater supply development and cost-effective waterwell construction (Foster 1984). In more limited areas, the weathering profile had been subsequently eroded, and only a fractured basement aquifer remained, which was much more patchy in distribution and much lower in exploitable groundwater storage. In parallel with (and slightly preceding) these efforts, UNICEF (with support from bilateral donors) mobilised low-cost, ultra-portable, lightweight drilling equipment to Peninsular India, with the intention of improving rural water-supplies in the innumerable villages of the droughtprone interior underlain by weathered granitic basement and Deccan Traps basalt (Black and Talbot 2005). The early outcome of these investments in groundwater understanding and development were: – Increased confidence in low-cost waterwell design amongst some groundwater engineers, and economical DOI 10.1007/s10040-011-0828-9

660

Fig. 1 Location maps for a Sub-Saharan Tropical Africa and b Peninsular India showing the general distribution of hard-rock aquifers

water supply of much improved microbiological quality and drought reliability in those rural areas where comprehensive coverage with hand-pump borewells or tubewells was successfully achieved. – Growth of a false impression amongst the rural development community that such waterwell supplies could be obtained virtually everywhere without hydrogeological Hydrogeology Journal (2012) 20: 659–672

studies and/or geophysical investigations, and be sited at ‘community convenience’ alone.

Relevance to the UN Millennium Development Goals In the subsequent decades, major investment effort (by national and state governments, multilateral and bilateral DOI 10.1007/s10040-011-0828-9

661

Fig. 2 Hydrogeological characteristics of the weathered crystalline basement in southeastern Africa (modified from Foster 1984)

donors, international charitable organizations, local communities or individual enterprise) has been put into trying to develop groundwater resources from weathered hardrock formations and this has resulted in a number of issues that require systematic scientific diagnosis, policy consideration and management action. For Tropical Africa, the issues can be briefly summarised as follows: – Escalating unit cost of village water-source provision in some areas, in significant part due to unacceptable groundwater chemical quality or inadequate yield. – Rapid growth in attempts to use groundwater in response to major increases in urban water-supply demand. – Consideration of possible increased use of waterwells of modest groundwater yield to support small-scale irrigated agriculture. In sharp contrast, Peninsular India initially experienced many years of steady improvement in the coverage of village water-supplies, facilitated by national industrialisation of low-cost waterwell technology. However, its subsequent (initially highly-successful) adaptation for irrigated agriculture, led to an ‘explosion’ of waterwell construction from the late 1980s (with millions having been drilled to depths steadily increasing from 30 to 60 m; Shah 2009) which resulted in: – Excessive resource exploitation on a widespread basis with lowering of water tables, complete drying-up of dugwells, increasing borewell failure and spiralling consumption of electrical energy (highly subsidised Hydrogeology Journal (2012) 20: 659–672

through politically-justifiable but fiscally crippling flatrate charging). – Reversal of the earlier trend towards achievement of universal provision of secure and safe sources of village water supply for the rural population, due to hydraulic interference and quality deterioration in drinking waterwells. – Massive private groundwater use in urban areas, raising numerous policy questions.

Consolidating scientific understanding Hard-rock aquifers of Sub-Saharan Tropical Africa

From the first conceptual model of weathered hard-rock aquifers (Figs. 2 and 3) it was evident that groundwater resource potential would be a complex interactive function of: – The type of hard-rock formation present. – Its structural condition defining primary jointing and fracturing. – Geomorphological evolution determining the depth of regolith weathering, degree of saprock fracturing and any subsequent erosion. – The present-day groundwater recharge and discharge regime controlling regolith saturated thickness and available drawdown to the upper part of the saprock. Experience in prospecting, developing and testing tubewells in the weathered crystalline basement in certain DOI 10.1007/s10040-011-0828-9

662

Fig. 3 a–b Typical groundwater occurrence and flow regime in weathered hard-rock aquifers

African countries (Malawi, Zimbabwe and Nigeria) indicated (Chilton and Foster 1995) that: – The aquifer can be tapped widely with low-cost handpump waterwells for rural water supply—in areas surveyed in detail, more than 80% of tubewells supported hand-pump yields (> 0.1 l/s) but less than 20% allowed motorised pumping (yield >1 l/s). – Groundwater production potential, however, is strictly limited by variable, but generally low, transmissivity (mainly in range 2–5 m2/d and only locally exceeding 10 m2/d). – Secondary permeability development has a complex relation with the weathering process, since while dissolution and leaching of minerals tend to increase porosity and permeability, the secondary minerals produced (such as kaolinite) tend to act in reverse, Hydrogeology Journal (2012) 20: 659–672

and while the more schistose metamorphic rocks tend to weather more readily, the presence of abundant biotite results in a low permeability regolith. – Despite uniform bedrock type, geomorphological situation and climatic regime, there are substantial (and unpredictable) local variations in waterwell yields and in aquifer response to pumping as a result of low regolith permeability (averaging<0.1 m/d) and generally poor connectivity of saprock fractures. – The upper part of the saprock provides most of the transmissivity, but the thickness of saturated saprolite above is important since it determines the ‘available drawdown’ and provides most of the aquifer storage (Sy usually>0.01); thus, the yield of most tubewells falls off dramatically with any dewatering of the saprock. – The development of larger water supplies for small towns or irrigated agriculture requires much greater DOI 10.1007/s10040-011-0828-9

663

investment to locate restricted zones where favourable lithology, structural features and deep weathering combine to provide higher transmissivity and available drawdown, and/or to design and construct waterwells of large effective radius. – While dispersed rural water-supply demands imply only very small average rates of groundwater abstraction (1–3 mm/a), and thus negligible pressure on the resource base, the development of larger water supplies always requires consideration of natural recharge rates (e.g. indications from parts of Malawi with annual rainfall of 800–900 mm/a are of recharge around 140 mm/a).

Hard-rock aquifers of Peninsular India Groundwater occurrence in the principal categories of hard-rock aquifer (weathered granitic basement and Deccan Traps basalts) shows some significant differences as a result of ‘pseudo-stratiform’ deposition of the latter, but essentially result in a four-fold hydrogeological division of the land area (Foster et al. 2007; Garduno et al. 2009; Limaye 2010) (Fig. 3): – A ‘run-off zone’ in which almost unweathered rocks occur virtually at the land surface and there is thus negligible groundwater storage (such areas still present major problems in respect of village water supply with continued dependence on expensive tankering or piped transfers). – A ‘recharge zone’ with sandy soil cover and fractured saprock below, but often with very limited groundwater storage capacity which may largely drain away in the dry season. – A ‘storage zone’ in which the regolith provides groundwater storage equivalent in volume to a few (usually 3–8) years of average recharge. – A ‘discharge zone’ of localised ‘down-gradient’ extent with reducing transmissivity and storage, but exhibiting natural groundwater discharge in the unpumped state— although widely natural discharge may have been eliminated due to heavy development for irrigation use. It should be noted that under some conditions of geological structure and geomorphological evolution the recharge zone and storage zone may in fact coincide, and the discharge zone can occur within the storage zone. Thus, within hard-rock formations, groundwater bodies with sufficient inter-annual storage to support major dryseason irrigation are not of universal distribution. The total available storage of these groundwater bodies is strictly limited by their weathering characteristics and water-bearing properties, and reduces markedly as the water table falls through the most productive saprock horizon (situated typically between 10 and 20 m below ground level (bgl)). Therefore, while investment in waterwell drilling in hard-rock terrains is always a gamble (in terms of obtaining a sustained yield sufficient for mechanised irrigation pumping), the risks increase and Hydrogeology Journal (2012) 20: 659–672

the success rate of new borewells drops markedly when excessive exploitation causes the water table to fall through the saprock horizon. In recent years detailed research—mainly in Andhra Pradesh (India) and latterly French-funded (e.g., PradeepRaj et al. 1996; Dewandel et al. 2006)—has provided further insights of the characteristics of the best developed parts of the weathered granitic basement aquifer (Fig. 3a; zone C): – The presence of at least two cycles of tropical weathering—one pre- and the other post-Tertiary uplift and erosion—which result in a regolith of higher permeability with pre-existing fractures preserved and a few metres of a more intensely fractured uppermost saprock. – This results in higher regolith permeability (0.1–2.0 m/d), which is correlated with residual fracture density, and much higher transmissivity values overall (usually >20 m2/d and sometimes >50 m2/d), and a storage coefficient of 0.02–0.03. – The groundwater body comprises two hydraulically interconnected layers—a phreatic layer in the weathered zone down to maximum depths of around 15 m and a deeper saprock horizon usually to about 30 m depth (exhibiting slight groundwater confinement in the valley floors). – About 90% of natural groundwater flow is usually concentrated in a 5 m or so thick horizon at the saprolith base and at the top of the fractured saprock, and thus the depth to the effective base of the aquifer is usually in the range 15–25 m. Over most drought-prone areas of Peninsular India, rainfall averages 500–800 mm/a and is concentrated in a single monsoon (June–August), during which potential natural recharge averages 70–100 mm/a. Understanding of recharge processes and their spatial variation is improving. Hill-slope and hill-front areas represent the most favourable areas for aquifer recharge (both natural and artificially enhanced) and it is here (Fig. 3a; zone B) that the aforementioned rainfall infiltration rates (aided by preferential flow) occur, whereas out across the pediplain rates fall-off to 30–40 mm/a. Most groundwater storage is of recent age and of low salinity, with just a small proportion entering much deeper fractures with aquifer residence times of 1,000+ years (Sukhija et al. 2005). In contrast, by the late 1990s, groundwater extraction rates had grown to reach an equivalent of 120–180 mm/a wherever useful groundwater bodies occur (curiously almost regardless of waterwell densities—given that the entire area is heavily populated and cultivated). As a consequence, the water table of most groundwater bodies declined steadily from the late 1980s (with only partial recovery in years of exceptional rainfall). The net decline of pre-monsoon water level has widely been 15–20 m over 25 years (Fig. 4), and in many areas the water table now stands almost permanently in the fractured saprock horizon. DOI 10.1007/s10040-011-0828-9

664 Fig. 4 Selected groundwater level hydrograph for the weathered granitic basement aquifer of Andhra Pradesh, India

Groundwater quality considerations Groundwater from weathered hard-rock aquifers is generally of good natural quality but exhibits substantial (and sometimes extreme) variations over short distances, both in depth and space, reflecting a locally complex, very sluggish, groundwater flow regime which results in ph/Eh favouring Fe solution and high Mg–SO4 concentrations, and (more seriously) elevated F concentrations. Fluoride (F) is beneficial for dental health, but optimal doses fall within a narrow range and the detrimental effects of chronic exposure to excessive fluoride include dental and skeletal fluorosis. Drinking water can be a major (and often the dominant) pathway of exposure to fluoride, and, thus, the WHO guideline value is 1.5 mg/l. High F concentrations in groundwater are geogenic (the result of water–rock interaction) and occur in Ca-poor/ Na–HCO3-dominant groundwaters, in which the solubility of fluorite (CaF2) is higher and dissolved F is stable at higher concentrations (Edmunds and Smedley 2005). Crystalline basement rocks, especially those of granitic composition, contain an abundance of F-bearing minerals, and the volcanic rocks of the East African Rift Valley also have high F concentrations. As a result, large populations in both Peninsular India and Sub-Saharan Africa are potentially exposed to fluorosis, and it is important to establish the local spatial and depth variations of fluoride in groundwater. The vulnerability of groundwater to pollution from surface anthropogenic activities is also greater than the generally low permeability suggests, because the vadose zone can be thin and preferential flow through saprolite macropores and cracks is ubiquitous (Chilton and Foster 1995).

Peninsular India: policy implications and management progress Balancing the cost-benefit of excessive rural groundwater exploitation First, it is necessary to consider what is meant by the term ‘excessive exploitation’ (often termed ‘overexploitation’) of groundwater resources and the dynamics of depletion of aquifer reserves. For most, the term ‘aquifer Hydrogeology Journal (2012) 20: 659–672

overexploitation’ implies a physically unsustainable situation in which the extraction of groundwater exceeds the replenishment, but in many weathered hard-rock aquifers, aquifer storage is very small and some parts of such aquifers can be fully replenished and virtually dewatered every year. In the end, the primary concern must be that the socioeconomic benefits of groundwater resource exploitation outweigh its direct and indirect (environmental) costs—the single most important factor being aquifer system susceptibility to irreversible degradation (through salinisation and/or compaction with land subsidence). This is not, however, the case for weathered hard-rock aquifers, with exception of the possibility of encountering increasing fluoride at depth in some areas. Superimposed on, and interacting with, the ‘physical picture’ of aquifers is the dynamic of groundwater demand— the primary initial driver for private groundwater use invariably being the unavailability or unreliability of surface-water supplies, and not resource availability or wellyield potential. The effect of these factors is then magnified by a series of secondary drivers including: – Flat-rate (or highly subsidised) electricity for irrigation well pumping (although in areas without electrification where more costly diesel fuel is used, excessive exploitation also occurs). – The low cost of borewell construction as a result of Indian manufacturing ingenuity. – Guaranteed prices for some crops with very high consumptive use of water such as paddy rice and sugar cane. The consequences of excessive exploitation and severe aquifer depletion are negative for the large majority of stakeholders (with the exception of waterwell drillers and pump suppliers), and can seriously prejudice efforts to alleviate rural poverty (Table 1). To correct a condition of serious groundwater resource imbalance, technical interventions which reduce groundwater demand and/or increase groundwater availability must be implemented: – Demand-side measures: interventions which (applied at field, village, district, state or federal level) have the effect of significantly reducing consumptive groundwater use through reduction of (1) the irrigated area, DOI 10.1007/s10040-011-0828-9

665

(2) the crop type, and with the same cropping, (3) nonbeneficial evaporation of irrigation water and (4) seepage of irrigation water to saline water bodies. – Groundwater recharge enhancement: physical engineering measures to retard and retain seasonal surface runoff and to provide conditions more conducive for infiltration to groundwater, with the objective of increasing the rainfall quantum stored in aquifers.

Spiralling consumption of electrical energy in irrigated agriculture Policy on rural electrification and electrical energy pricing has exerted a great influence on groundwater use for irrigated agriculture in Peninsular India (World Bank 2010). In many areas, there is a high coverage of rural electrification and the supply (although very intermittent) is sufficiently predictable overall for farmers to rely exclusively on electric-engine pumpsets for lifting groundwater, whilst in others they have diesel engines as a standby. The major electricity subsidy via flat-rate tariffs, related only to pump horse-power, results in farmers actually paying for less than 20% of the energy supplied, and insulates them to a degree from the cost of falling water tables. It is unlikely that flat-rate electrical tariffs are the primary cause of excessive groundwater exploitation (because this condition also widely occurs where farmers have to use diesel-engined pumpsets), but the existence of flat-rate electricity tariffs where shallow low-storage hardrock aquifers predominate has allowed extremely inefficient practices to develop such as: – Farmers leaving pumps switched on to obtain supply when the power-system activates (since this is not continuous or regular) and not worrying about energy losses. – Farmers continuing to operate borewells and pumps at groundwater levels which are far too low and at which well entry and pump friction losses are very high (Fig. 5), and this marginal extraction would be completely uneconomic if farmers felt the full cost of electrical energy consumed.

The main effect of the flat-rate electrical energy tariffs in hard-rock aquifer terrains has thus been to impose an enormous financial burden for little return on state electrical energy generation and distribution companies. It is also evident that the large electrical energy subsidy is not benefiting the poorest farmers—because their waterwells are usually less deep and often dry-up early in the rabi (cold dry) season—and it goes to the somewhat better-off farmers who have deeper borewells (and whose groundwater production is probably the more inefficient; Shah 2009). Paradoxically, in political terms, some form of rural energy subsidy (or other allowance) is justified (World Bank 2010) to address the iniquitous price differential for irrigation water supply between farmers outside irrigation canal commands (who are wholly dependent on groundwater) compared to those within command areas (where traditionally the hydraulic infrastructure has been provided free by the government and where irrigation water pricing recovers only a small fraction of operation and maintenance costs). Indeed the policy might be manageable in technico-economic terms in higher-yielding aquifers (for which unit pumping energy requirements and waterwell energy losses are much smaller), but it is an extremely costly policy in hard-rock aquifer terrains. Thus, rural electricity use-efficiency mapping is urgently needed to inform the electrical energy-groundwater policy debate and understand better the current patterns of rural use and grid distribution losses according to aquifer type and groundwater levels. Viewed from the groundwater perspective, the overriding need is to find a way of facilitating recovery of watertable levels, such that the most productive horizon of weathered hard-rock aquifers remains partially saturated in the dry season. The benefits of such an outcome, however achieved, would be a major reduction in electricity consumption for only modest reductions in irrigation water-supply availability (Shah 2009).

Unrealistic expectations for managed recharge enhancement Over the past decade or so, the predominant policy response to indications of excessive groundwater

Table 1 Implications of groundwater overexploitation in hard-rock aquifers for rural poverty alleviation Consequences of excessive exploitation

Implications for poverty alleviation

Higher waterwell construction costs Competitive waterwell deepening Decreasing waterwell yields Failed waterwell investments

Larger drilling depths/costs for (poorer) ‘late-comers’, who usually have worst-situated land holdings (with less chance of successful waterwells) Richer farmers are more able to finance waterwell deepening and poorer farmers less able to obtain and repay bank loans Crop loss or poor unit crop yields, reduced cultivated/irrigated area, serious reductions in farmer incomes Loan repayment default with spiralling debt, sale of land at low prices, discredit socially and even suicide, increased exposure for ‘rural development banks’ resulting in more expensive loans prejudicing poorer farmers Failure of drinking-water sources and quality deterioration (F content or salinity) resulting in increased water-collection distances and return to ‘unsafe sources’

Village water-source failure

Hydrogeology Journal (2012) 20: 659–672

DOI 10.1007/s10040-011-0828-9

666

Fig. 5 Hydrogeological profile of an Indian hard-rock aquifer with indicative variation of pumping yields and energy costs with falling water table (gwl)

abstraction for agricultural irrigation from the weathered hard-rock aquifers has been to promote aquifer recharge enhancement, in association with the implementation of rainwater harvesting techniques. Championed initially by advocates of watershed development and community stewardship of natural resources, artificial groundwater recharge has gathered momentum both within and outside government (with Union and States espousing the concept in a major way) and there now exists a veritable ‘groundwater recharge movement’ in India (Sakthivadivel 2007) backed by India’s Groundwater Recharge Master Plan. However, this follows two criteria for identifying recharge—availability of surplus run-off and availability of storage space in aquifers—which result in investments being driven by the potential for groundwater recharge enhancement and not by the need for additional resources. A large proportion of the over-exploited groundwater blocks in India exist in semi-arid areas underlain by hard-rock aquifers (in Andhra Pradesh, Karnataka, Madhya Pradesh, Maharashtra and Tamil Nadu), where the quantum of surface run-off for potential recharge is limited and arises over a short period, and only restricted parts of the land area have the most favourable hydrogeological conditions for recharge structures. Moreover, monitoring of various recharge structures suggests that the incremental increase in recharge rate is usually less than 10% (Gale et al. 2006; Dillon et al. 2009). Thus, it is now becoming accepted that the success of investments in recharge enhancement (totalling more than US$ 1,000 million in three major programs) has been very patchy—although it can be argued that inadequate technical backstopping and poor-quality implementation have been a significant factor behind the mixed results. Hydrogeology Journal (2012) 20: 659–672

When compared with the scale of overdevelopment of many threatened groundwater blocks, it becomes clear that, while cost-effective groundwater recharge in these areas is useful, it will not come close to bridging the gap between supply and demand (Table 2). Recharge enhancement in these areas, therefore, does not offer an alternative solution (World Bank 2010), but a valuable complement which should ideally be reserved for priority local uses such as increasing drinking-water source security, or as incentive for community-based groundwater management initiatives.

Mobilising community-based management of groundwater demand In Peninsular India, approaches to groundwater management must take account of the following: – The extremely large number of individually small users – Limited institutional capacity for resource management (needing to be focused on the few critical aquifers of major potential which are at risk of irreversible degradation). – Characteristics of the predominant hard-rock aquifers, which mean that pumping drawdown effects are localised (restricted to the immediate micro-watershed and in many cases village boundaries)-ignoring for the moment the effects of diffuse stream baseflow diminution. – The need to ensure that water-table lowering will not be accompanied by irreversible aquifer side-effects and/or environmental degradation (although troublesome fluoride concentrations may arise). Thus, community self-regulation of groundwater use is favoured as the most realistic option—there being DOI 10.1007/s10040-011-0828-9

667 Table 2 Qualitative assessment of groundwater recharge enhancement measures

important examples of this approach in both Maharashtra (Foster et al. 2009) and Andhra Pradesh (Garduno et al. 2009). Hivre Bazar is a village (of 1,200 population and 975 ha area) in the elevated drought-prone Deccan Traps country of Maharashtra (450 mm/a average rainfall)—here the weathered zone of the Deccan Traps basalt reaches to about 12–15 m bgl and is underlain by massive (sparingly fractured) basalt providing only very limited additional groundwater flow. Under the leadership of an informed and charismatic village council chief, a concerted effort on groundwater management commenced in 1994 (as part of the Maharashtra Ideal Village Social Development Scheme) with implementation of a comprehensive 5year plan (Foster et al. 2009), following a long history of drought propensity and land degradation, with farmers struggling to feed their families and cattle without leaving the village periodically to search for paid work. In Hivre Bazar, staple hard-grain crops are grown primarily for home consumption with residues serving as livestock fodder or domestic fuel, while most pulses, onions, vegetables and flowers are sold at market. In the most favourable years, almost 60% of the land can be irrigated, but in drought years rabi (cold dry season) and jayaad (hot dry season) crops have to be radically reduced. During the 1990s, the main groundwater-related decisions made by the village council (on its chief’s advice) comprised: – Most critically, prohibiting the use of borewells (and the drilling of vertical bores in dugwells) for agricultural irrigation, which had the great benefit of moving farmers’ minds and resources away from competition for deeper groundwater to cooperation on maximising benefits from groundwater to which they nearly all had access. – Subjecting the micro-watershed to comprehensive reforestation and water harvesting—notably hill contour-trenching, Nalla stream bunds, prohibiting axe use (dung replacing timber for domestic heating) and a livestock grazing ban (with scythes hired to hand-cut fodder for animal stall feeding), which is estimated to have increased local groundwater recharge rates from around 70 to 100 mm/a. Hydrogeology Journal (2012) 20: 659–672

– Banning sugar-cane cultivation (given its high wateruse and other implications). Additionally and importantly, village-level crop-water budgets were introduced in 2002. The post-monsoon availability of soil-water and groundwater are estimated from field data, human and livestock water needs given first priority, and then (using past experience) the availability for irrigated cultivation is calculated and compared to the aggregate need of villagers’ proposed cropping. In dry years, villagers are asked to reduce their proposed irrigated area and give preference to low-water demand crops, with mutual surveillance usually being sufficient to achieve compliance. Such proactive resource management has resulted in a marked contrast between Hivre Bazar and surrounding villages. In the jayaad (hot dry) season, as many as 32 dugwells produce important revenue from irrigated onion, vegetable and flower cultivation, and only a few dugwells in the upper watershed dried out. The household-level benefits of community land and water management include average family incomes rising markedly (to over US$ 500/a) and land-values appreciating many-fold over 15 years. Andhra Pradesh is mainly underlain by the weathered granitic basement rocks and receives an average rainfall of 650–950 mm/a, from which natural recharge rates are believed to average 70–100 mm/a. In contrast, by the mid1990s, groundwater extraction rates had grown to an equivalent of 120–150 mm/a (Pradeep-Raj et al. 1996). Groundwater is exploited by dugwells penetrating to just below the weathered zone and borewells mainly from 30 to 50 m deep (of variable yield but with 60% achieving >2 l/s). During the past 30 years, the number of dugwells has remained at about 0.9 million but, with an increasingly large portion falling dry or becoming ‘seasonal’, there has been rapid growth in the number of borewells to the current estimate of 1.7 million (with average yields decreasing and depths increasing) and widespread excessive exploitation of available resources with serious dewatering of the main water-bearing horizons of the shallow aquifer system. The (Indian) Central Groundwater Board (CGWB) groundwater resource estimation methodology (not ideal for this type of hydrogeologic terrain) was last used to assess groundwater resource status in Andhra DOI 10.1007/s10040-011-0828-9

668

Pradesh in 2008, and it suggested that of the 1,227 blocks (of 100–300 km2) into which the state is divided, 508 were at a critical or semi-critical state of development (World Bank 2010). Additionally, when only non-irrigation canal command blocks were considered, the figure rises to over 80% (even before considering the need to allow for minimal downstream needs). The pioneering AP-WELL Programme of the 1990s, supported by the Food and Agricultural Organisation of the UN (FAO-UN), covered some 14,500 marginal farmers using 14,000 ha of irrigated land in 370 villages in most of drought-prone Andhra Pradesh. Without offering any cash incentives or subsidies, the programme developed participatory hydrological monitoring to provide farmers with the necessary knowledge, data and skills to understand groundwater resources and to manage their use through controlling on-farm demand for water (Garduno et al. 2009). The subsequent AP-FAMGS Programme (with UN-FAO and bilateral funding and World Bank support on evaluation), which commenced in 2007, makes the strong link between groundwater availability and irrigation use but leaves farmers free to make crop planting decisions and extract groundwater as they desire. Nevertheless, in a majority of pilot project areas, the results have been very positive (Garduno et al. 2009) as witnessed by:

efficiency through the introduction of modern precision irrigation techniques will solve the problem of groundwater over-exploitation. It has to be recognised that with current irrigation practices, which predominantly involve traditional small-scale flood irrigation, a substantial proportion of the total water losses actually takes the form of infiltration and a significant proportion of the applied irrigation water returns to groundwater; thus, reducing total soil-water losses does not necessarily represent a real water-resource saving. Other practical measures that could be considered by state governments in cooperation with community groups to confront excessive groundwater exploitation are:

cation and irrigation water-saving techniques—with 42% of areas consistently reducing the groundwater overdraft over 3 years and a further 51% achieving intermittent reductions. – Farmers increasing profitability while using less water— with reduction in groundwater overdraft coming from ‘multiple individual management decisions’ rather than ‘altruistic collective action’.

Detailed evaluation of groundwater use from weathered hard-rock aquifers in urban areas is restricted to Aurangabad City in Maharashtra and Bangalore in Karnataka (Foster and Mandavkar 2008; Gronwall et al. 2010), but general indications from various other cities suggest these are in many ways typical. In all cases, private waterwell construction and dependency on in situ self-supply from groundwater has mushroomed in recent years as a ‘coping strategy’, given the widely inadequate service levels of municipal water supply. In most wards of Aurangabad City, the municipal water supply provides less than 1-in-24-h service at low mains pressure. In order to reduce dependence on the purchase of much more expensive tanker water supply (at US$ 1.0 +/m3), urban dwellers and commercial water users in particular have widely turned to private borewell construction as an alternative source. This has proved to be very effective from the perspective of generating additional water supply for users (Fig. 6), despite the low yield potential and very limited reserves of the weathered Deccan Traps basalt underlying the city. It also demonstrates an implicit capacity to pay for more reliable urban water supply, since the capital and running costs of private groundwater supplies are significantly above those of the current highly subsidised domestic urban water tariff. However, this major urban private investment in ‘selfsupply’ from shallow groundwater is not without its problems. In areas of very deficient municipal water supply and higher population density, it usually means heavy seasonal depletion (almost emptying) of the lowstorage groundwater body with widespread borewell failure (especially in May–June). In addition, concerns

– Introducing some form of rationing of (more reliable) rural electrical power-supply to village communities commensurate with sustainable use of groundwater resources and more efficient use of electricity supplies (Shah and Verma 2008; World Bank 2010). – Locally agreed protocols for minimum distances between irrigation waterwells and/or ‘drilling prohibition zones’ around village drinking-water sources.

Defining a balanced groundwater use policy – Reduction in groundwater use through crop diversifi- for urban areas

The up-scaling and replication of this very positive experience will necessitate a flexible phased approach which engages experienced support organizations, together with development of a ‘lighthouse function’ in the state groundwater department to monitor the process and to ensure continuity and momentum. The fact that community-based approaches are likely to offer the most effective and pragmatic means for groundwater management does not (and must not) diminish the responsibility of state government agencies to take action, which needs to include: – Provision of transparent information on resource status. – Extension support for cropping plans to reduce groundwater and energy use, and concomitantly increase crop productivity per unit energy and water consumption. – Reform or realignment of policies in other sectors acting as drivers of groundwater use, including looking at more creative ways for targeting the electrical energy subsidy and providing incentives for reductions in electricity and groundwater use. In this context, state governments will need to resist the temptation to think that simply improving irrigation water Hydrogeology Journal (2012) 20: 659–672

DOI 10.1007/s10040-011-0828-9

669

are growing about bacteriological and chemical pollution, and nutrient overloading of aquifers. However, perhaps the greatest concern is that in Aurangabad, and in other cities on weathered hard-rock aquifers, the existing private access at moderate cost to in-situ groundwater will inevitably be a key factor affecting the cost-recovery potential for major new urban water-supply schemes based on expensive transport and treatment of water from distant surface-water sources (whose full cost including the recovery of capital investment will be much higher than the cost of private groundwater supplies). It is clear that in such cities, a much more integrated vision of, and balanced policy between, utility infrastructure provision (both water supply and sanitation) and private self-supply will need to be developed on a case-by-case basis, including direct measures to improve the availability and reduce the quality risk associated with the use of in situ groundwater.

Some exceptions in hard-rock terrain needing a different approach

Over most of ‘hard-rock India’, the use of groundwater for irrigation is largely by subsistence farmers, who cultivate primarily to provide for their own family needs and sell only modest excess production and some special crops in local markets. However, it should be noted that there are occasional small but important groundwater bodies, in coastal or graben-filled sedimentary aquifers (such as those of the Tapi Graben in Maharashtra and the Kortalaiyar Basin in Tamil Nadu), which offer much greater potential in terms of waterwell yields and storage reserves. Their groundwater is often under exploitation for commercial agriculture with production of cash crops (such as banana plantations and grape cultivation) for major national markets and for export, and also sometimes for large-scale abstraction for the urban water supply of municipal utilities. Due to the combination of very intense

resource exploitation and intrinsic susceptibility of the aquifer system, these groundwater bodies may be under threat of irreversible degradation, for example from saline intrusion. It follows that where groundwater bodies under exploitation for large-scale agro-industrial production and/or municipal water supply are at risk of irreversible degradation, there is a strong case for a greater element of state resource regulation and investment in engineered recharge to provide a solid framework in which demand-management measures can be required of commercial groundwater users. Since the total consumptive water requirement of highly profitable fruit cultivation (even when using efficient systems of drip irrigation) can be around 1,500 mm/a, achieving groundwater use sustainability in these localised high-potential but susceptible aquifer systems will require a major long-term effort including: – ‘Legal notification’ that the aquifer is at risk of irreversible degradation, requiring the State government to register all existing operational waterwells by a specific date (and declare any remaining unregistered illegal), to prohibit construction or deepening of irrigation tubewells, and to report progress on the management measures to the Union Parliament. – Enforcing an ‘overall ceiling’ on irrigation use through a system of individual and community-aggregated groundwater rights and thereby reducing the proportion of cultivated land area under irrigation. – Promoting and financing the use of excess wet-season canal flows for aquifer recharge via disused dugwells.

Sub-Saharan Tropical Africa: policy implications and management needs The evidence of the present overview strongly suggests that the yield potential of weathered hard-rock aquifers is, in

Fig. 6 a Waterwell costs and b annual water use by various user categories in Aurangabad City, India Hydrogeology Journal (2012) 20: 659–672

DOI 10.1007/s10040-011-0828-9

670

general, notably lower in Tropical Africa than in Peninsular India. This means that the possibility of obtaining waterwell yields sufficient for installation of motorised pumps is much more locally developed, within certain areas of structural depression (in which the weathering zone is deeper and overlain by coalluvial deposits). Although more systematic hydrogeological surveying is required to confirm this conclusion, should it prove to be the case, it will have farreaching policy implications about the extent to which this extensive aquifer system (across southeastern, central and western Africa) can be developed for urban water supply and irrigated agriculture.

Improving the efficiency of rural village water-supply provision It is essential that groundwater resources of the weathered crystalline basement be developed further to provide adequate rural water-supply coverage, since at present this only stands at about 40% overall in Tropical Africa. Groundwater remains the only potentially viable option for improved rural water supply over extensive areas; however, currently, development programmes serving the basic health and livelihood needs of rural communities find it difficult to support capital costs in excess of US$ 3,000/waterwell, implying an absolute need (Foster et al. 2008) to: – Keep costs down by constructing shallow waterwells with appropriate completion, and selecting low-cost drilling techniques. – Plan, procure and implement rural water-supply work in such a way as to permit ‘rolling projects’, guided by sound hydrogeological and hydrogeochemical advice, and using appropriate hydrogeophysical equipment, and learning as they progress so as to improve efficiency and reduce cost. There remain many instances of ‘drilling blind’ in rural water-supply provision, constructing unnecessarily deep waterwells, and failing potable drinking-water quality standards (due to hazardous or troublesome constituents such as fluoride, soluble iron or salinity; Chilton and Foster 1995; Edmunds and Smedley 2005; Smedley et al. 2007). Such problems are widely increasing the unit cost of rural water-supply provision, and in some areas mean that adequate sources are not being provided. The only way to overcome them is to use scientific information and equipment to guide waterwell siting and design, incorporating it in a cost-effective and integrated fashion.

Assessing prospects and controlling hazards of urban groundwater use Rapid growth of urban population (widely at 2–7%/a) and water demand (up to 10%/a) is a reality in Tropical Africa, and will be further accentuated in some climate-change scenarios. This is not just occurring in mega-cities (including those situated on weathered basement rocks Hydrogeology Journal (2012) 20: 659–672

such as Harare in Zimbabwe), but also hundreds of medium-sized towns (Foster 2009), and is highly relevant to water supply of the poorest dwellers (Gronwall et al. 2010). Where suitable aquifers are present (including the weathered crystalline basement, at least in zones where it offers best yield potential) expansion of groundwater use will be the preferred response—in terms of time taken, capital outlay and drought reliability. In urban areas located over weathered crystalline basement, the challenge will be to locate zones that allow construction of waterwells with sufficient yield and assured quality to support continuous pumping and reticulation to standposts. Thus, priority must be put on improving groundwater resource and vulnerability appraisal, efficient waterwell design, aquifer recharge and wellhead protection, together with improved monitoring as a sound basis for future expansion. Given the paucity of reliable data, it is far from a trivial task to establish the present level and current trends in groundwater use. The World Bank–Africa Infrastructure Country Diagnostic (which uses a 2007 database incorporating 63 large-scale surveys in 30 countries) reveals substantial variation between more and less urbanized countries; however, the following conclusions (Foster 2009) can be reached: – On average only 38% of urban dwellers are served by mains water supply piped to their dwelling, although a further 29% have municipal stand-posts within 500 m (some supplied by groundwater). – 24% of urban water supply (by user numbers not volume of abstraction) is groundwater directly collected from waterwells constructed by municipal, community or private initiative—and this is the most rapidlygrowing category at 1.5 %/a on average and over 5%/a in some countries. – On average, 33% of urban populations are served by waterborne sewerage, but 59% are dependent upon in situ sanitation (mainly pit latrines with some septic tanks) and about 8% have no sanitation system whatsoever. Contrary to operational recommendations, only a minor proportion of pit latrines are periodically emptied (most being connected to supplementary pits), implying the accumulation of a large contaminant load to groundwater in shallow aquifers (especially in highly populated areas). The significance of pit-latrine sanitation for groundwater quality is well established (Xu and Usher 2006), with a high-nitrate problem coupled with other chemical and fecal contamination having been identified in Mochudi in Botswana almost 30 years ago (Lewis et al. 1980). The risk of fecal pollution, however, could be limited to the most vulnerable hydrogeological conditions, but it currently remains a much more widespread problem because of inappropriate sanitation unit design/operation and inadequate waterwell sanitary completion. The most-probable future trend will be expansion of lowcost facilities such as borewells and improved pit latrines for DOI 10.1007/s10040-011-0828-9

671

sanitation (Foster 2009). However, in Tropical Africa, waterwell construction costs still remain high—for example, small-diameter shallow tubewells with simple plastic casing equipped with handpumps cost US$ 3,000–5,000 (supplying say 200 people at US$ 15/capita capital investment). Urban waterwell construction is likely to continue mainly under national, state or municipal funding, together with support from international donors and charities. The demand for selfsupply from groundwater by residential, commercial and industrial users is also likely to grow substantially; thus, pragmatic ways need to be found for ‘living with’ urban quality deterioration problems through: – Providing incentives for logical types of private urban groundwater use (such as domestic toilet flushing, laundry, amenity irrigation, non-sensitive industries, cooling water, etc.). – Being aware of potential long-term operational and financial problems created by large-scale residential in situ self-supply, and the potential public health hazard in highly vulnerable aquifers. – Considering measures to reduce subsurface contaminant load (especially regular emptying of existing in situ sanitation facilities, introducing dry or eco-sanitation units, prioritizing mains sewerage in areas of high aquifer pollution vulnerability and/or industrial effluent generation). – Enhancing aquifer recharge by rain-water harvesting from roof-top and paved areas. In all cases, more effort needs to go into defining appropriate policies for urban groundwater use and on evaluating, managing, conserving and protecting resources. In fast-growing urban centres, this will normally require an integrated effort, involving a consortium (or standing committee) of empowered representatives from the water-resource regulator (where such exists), the water utility, the public-health authority, the municipal land-use agency and civil society organisations, with a mechanism for community consultation and a technical support group to evaluate specific issues and potential conflicts. Professional expertise on how to evaluate, develop, manage and protect groundwater resources has declined in some parts of Sub-Saharan Africa since the 1980s (especially in government offices) and the human resource dimension of ‘making better use of urban groundwater resources’ cannot be overlooked.

Appropriate use of groundwater potential for agricultural irrigation The more traditional use of groundwater for irrigated cultivation is restricted to village garden-scale cultivation of vegetables using hand-pumps and manual watering (Foster 1984); and this certainly should be encouraged and amplified since it helps to improve family nutrition, either by constructing stand-alone waterwells (Barry et al. 2010) and/or using surplus yield after drinking-water requirements are satisfied. Hydrogeology Journal (2012) 20: 659–672

Groundwater irrigation from waterwells with motorised pumps has only been very sparsely developed to date in Tropical Africa in general. And, in the case of the weathered crystalline basement aquifer in particular, it has to be questioned whether adequate waterwell yield potential exists on a widespread basis. If major investment programmes are embarked upon without detailed hydrogeological feasibility studies, there will be a high risk of failure due to the low success rate of drilling and/or nonsustainability of well yields. Nevertheless, the potential for successful and sustainable development in some (probably limited) areas undoubtedly exists—and it is strongly recommended that, prior to investment in large-scale waterwell construction and rural electrification, careful hydrogeological investigation is undertaken to guide phased, monitored and evaluated pilot development, with the philosophy of ‘learning as you go’. In this context much can be learnt from the initially successful development of groundwater irrigation development in Peninsular India, whilst avoiding the problems that have occurred there through excessive exploitation, unsustainable investment and chaos in the electricalenergy sector as a result of indiscriminate blanket rural energy subsidies. Acknowledgements The author is most grateful to Dr S.N. Rai (National Geophysical Research Institute of India) and other members of the organizing committee of the Joint IAH and IAHS Joint Congress of Hyderabad, India (September 2009), for the invitation to give this paper, in the capacity of IAH Immediate Past President, as the IAH Keynote Lecture to that Congress. This came exactly 25years after an earlier (similar) invitation (for Tropical Africa only) at the IAHS African Symposium of Harare, Zimbabwe, in 1984. The author wishes to acknowledge valuable discussions with his former colleagues in the following organizations: World Bank, Delhi Office (Sanjay Pahuja, Smita Misra and R.S. Pathak), GW-MATE (Hector Garduno, Albert Tuinhof and Frank van Steenbergen) and British Geological Survey (John Chilton, Alan Macdonald, Adrian Lawrence, Jeff Davies and Pauline Smedley), from which he developed some of the opinions expressed in this paper. In addition, he valued interaction with colleagues in India (L. Pradeep-Raj of the Andhra Pradesh-Groundwater Department, Suresh Khandale of the Maharashtra Groundwater Surveys and Development Agency and Tushaar Shah of the International Water Management Institute), in Africa (Lister Kongola, Callist Tindunimugaya, Segun Adelana and John Farr of the IAH Burdon Network) and Jacob Burke of the UN Food and Agriculture Organization in Rome.

References Barry B, Kortatsi B, Forkour G, Krishna-Gumma M, Namara R, Rebelo L-M, Berg J van den, Laube W (2010) Shallow groundwater in the Atankwidi catchment of the White Volta Basin: current status and future sustainability. IWMI research report 139, IWMI, Colombo, Sri Lanka Black M, Talbot R (2005) Water: a matter of life and health—watersupply and sanitation in village India. UNICEF, New York; Oxford University Press, Oxford Chilton PJ, Foster SSD (1995) Hydrogeological characterization and water-supply potential of basement aquifers in Tropical Africa. Hydrogeol J 3:36–49 Chilton PJ, Smith-Carington AK (1984) Characteristics of weathered basement aquifer in Malawi in relation to rural watersupplies. IAHS Publ. 144, Wallingford, UK, pp 57–72 DOI 10.1007/s10040-011-0828-9

672 Dewandel B, Lachassagne P, Wyns R, Marechal JC, Krishnamurthy NS (2006) A generalised 3-dimensional geological and hydrogeological conceptual model of granite aquifers controlled by single or multiphase weathering. J Hydrol 330:260–284 Dillon P, Gale I, Contreras S, Pavelic P, Evans R, Ward J (2009) Managing aquifer recharge and discharge to sustain irrigation livelihoods under water scarcity and climate change. IAHS Publ. 330, IAHS, Wallingford, UK, pp1–12 Edmunds WM, Smedley PL (2005) Fluoride in natural waters. In: Essentials of medical geology, chap 12. Elsevier, Amsterdam, pp 301–329 Foster S (1984) African groundwater development: the challenges for hydrological science. IAHS Publ. 144, IAHS, Wallingford, UK, pp 3–12 Foster S (2009) Urban water-supply security in Sub-Saharan Africa: making best use of groundwater. IAHS Publ. 334, IAHS, Wallingford, UK, pp221–228 Foster S, Mandavkar Y (2008) Groundwater use in Aurangabad: a survey and analysis of social significance and policy implications for a medium-sized India city. World Bank GW-MATE Case Profile Collection 21. http://www.worldbank.org/gwmate. Cited 15 June 2011 Foster S, Garduno H, Tuinhof A (2007) Confronting the groundwater management challenge in the Deccan Traps country of Maharashtra-India. World Bank GW-MATE Case Profile Collection 18. http://www.worldbank.org/gwmate. Cited 15 June 2011 Foster S, Tuinhof A, Garduno H (2008) Groundwater in SubSaharan Africa: a strategic overview of developmental issues. IAH Selected Papers in Hydrogeology, vol 13, Taylor and Francis, London, pp 9–21 Foster S, Limaye S, Mandavkar Y, Msangi S (2009) A hydrogeologic and socioeconomic evaluation of community-based groundwater resource management: the case of Hivre Bazaar in Maharashtra-India. World Bank GW-MATE Case Profile Collection 22. http://www.worldbank.org/gwmate. Cited 15 June 2011 Gale IN, Macdonald DMJ, Calow RC, Neumann I, Moench M, Kulkarni H, Mudrakartha S, Palanisami K (2006) Managed aquifer recharge: an assessment of its role and effectiveness in watershed management. British Geological Survey commissioned report CR/06/107N, BGS, Keyworth, UK

Hydrogeology Journal (2012) 20: 659–672

Garduno H, Foster S, Pradeep-Raj L, Steenberger F van (2009) Addressing groundwater depletion through community-based management actions in the weathered granitic basement aquifer of drought-prone Andhra Pradesh-India. World Bank GWMATE Case Profile Collection 19. http://www.worldbank.org/ gwmate. Cited 15 June 2011 Gronwall JT, Mulenga M, McGranahan G (2010) Groundwater selfsupply and poor urban dwellers: a review with case studies of Bangalore and Lusaka. IIED Human Settlements Programme Publication, IIED, London Lewis WJ, Farr JL, Foster SSD (1980) The pollution hazard to village water-supplies in eastern Botswana. Proc Inst Civ Eng (London) II 69:281–294 Limaye SD (2010) Groundwater development and management in the Deccan Traps (basalts) of western India. Hydrogeol J 18:543–558 Pradeep-Raj L, Nandulaland G, Soni K (1996) Nature of aquifer in parts of granitic terrain in Mahabubnagar District, Andhra Pradesh. J Geol Soc India 48:299–307 Sakthivadivel R (2007) The groundwater recharge movement in India. In: Comprehensive assessment of water management. CABI, Agriculture series, CABI, Wallingford, UK Shah T (2009) Taming the anarchy: groundwater governance in India. Resources for the Future, Washington, DC Shah T, Verma S (2008) Real-time co-management of electricity and groundwater: an assessment of Gujurat’s pioneering Jyotirgram Scheme. Econ Polit Wkly (New Delhi-India) 43:59–66 Smedley PL, Knudsen J, Maiga D (2007) Arsenic in groundwater from mineralised Proterozoic basement rocks of Burkina Faso. Appl Geochem 22:1074–1092 Sukhija BS, Reddy DV, Nagabhushanam P, Bhattacharya SK, Jani RA, Kumar D (2005) Characterisation of recharge processes and groundwater flow mechanisms in weathered-fractured granites of Hyderabad (India) using isotopes. Hydrogeol J 14:663–674 World Bank (2010) Deep wells and prudence: towards pragmatic action for addressing groundwater overexploitation in India. World Bank, Washington, DC Xu Y, Usher B (2006) Groundwater pollution in Africa. Balkema Proceedings and Monographs in Engineering, Water and Earth Sciences, Taylor and Francis, London

DOI 10.1007/s10040-011-0828-9