WATER DEVELOPMENT PROJECT
ECAFE (1971) examined the cost of water development based on a number of projects that
had been funded by the United Nations in Asia. The following discussion is based on their
report as it relates to the capital and operation and maintenance costs associated with
tubewells and the dependency of these costs on discount rates and aquifer
transmissivity. Their information has been supplemented with the experience gained
by AGW Consultants in its projects in Africa.
A local or national plan for water development is the outcome of a two-way information
exchange between planners and beneficiaries. A development plan is more than a list
of desirable goals; it is a projection of progress toward meeting objectives, establishing
specific targets and a plan to achieve them. Water development plans do not
necessarily consider high rates of return on capital as highly as meeting targeted
results. That is, the actual capital costs commonly are secondary to the public
However, where capital resources are scarce, a maximum return on investment becomes
almost as important as achieving targeted objectives whether it be the supply of municipal
water for towns, villages or cities, or the supply of water for agricultural
projects. Other benefits from water development plans such as increased employment
and increased generation of taxes is not generally considered when evaluating the
economics of such plans for acceptance.
It is important at each step in a project to evaluate alternative means for solving
technical problems to meet targeted objectives. Items normally examined may be the
relationship between depth of water and returns from the crop in an agricultural
project. In this case, the optimum depth to water is obtained by comparing the
additional cost of lowering drains and to additional output with returns to other
A major technical issue that has almost never been considered is the increased cost to
a project, both in terms of capital investment and long term operation and maintenance
costs, of locating tubewells in areas of low aquifer transmissivity or of locating wells
with less than an acceptable capacity. Water planners in Africa generally establish
10 liters per hour (l/hr) as an acceptable well yield.
Transmissivity is a number that describes the rate at which ground water flows through
an aquifer under the influence of a unit hydraulic gradient. Well yield is directly
proportional to aquifer transmissivity. For two wells of identical construction
detail and production capacity, the overall costs of construction and operation and
maintenance will be higher for the tubewell in the zone of lower transmissivity. The
difference in overall cost increases with higher ground-water production rates.
Clearly, any water development plan will require a multidisciplinary approach including
engineering, geology, soil chemistry, agriculture, sociology, and finance. Each of
these disciplines requires specialized data to make informed choices.
The most serious hazard in data collection is the temptation to collect inappropriate
data or too much data and to leave too little time for data analysis. Each type of
investigation should be based on a clear understanding the type of data necessary and why
it is required.
The "shotgun" approach of collecting masses of data and conducting unfocused
studies with the hope that the data may prove important cannot be supported.
The tendency for project personnel to conduct studies in their own areas of expertise
because that is what they do cannot be supported either.
Researchers estimate that it can take nine times as long to analyze the data as to
collect it. Therefore, irrelevant data should not be collected because it wastes
analytical personnel. We should only collect data that will lead to precise
conclusions at the earliest moment in a project. We believe focused investigations
will make other aspects of a project unnecessary.
PROJECT DESIGN CRITERIA
Regardless of project location, capital is generally the scarcest resource. It
must be allocated efficiently between competing alternative investments.
Commonly, financing water development projects encourages an over-capitalized
design. It is relatively easy to obtain capital for original construction through
bond issues or from a central government or an international financing agency. But,
it is much more difficult, if not impossible, to obtain financing for recurring operation
and maintenance expenditures. Yet, the capitalized value of savings in capital
expenditure will finance many times the cost of delayed recurrent expenditure.
This is especially true in developing countries, where capital scarcity and a wide
range of profitable investment opportunities result in high discount rates or a high value
placed on current capital availability.
We must consider two elements in designing a ground-water development project using
tubewells. First, water is a stock resource, which has been stored slowly in an
earthen reservoir over a long period of time. Second, ground water may be a
renewable resource from seepage from rivers and canals and percolation from fields after
rainfall or irrigation.
If recharge is negligible, the only option open is the irreversible exploitation
or "mining" of ground water. Such regimes exist at Kufra and Sarir in
Libya where the Nubian Sandstone is being mined of its water. In the Chad Basin
artesian wells no longer flow. The Ogallala Formation of the High Plains in the
United States is being significantly dewatered. The aquifers below the Cities of
Venice and Mexico City have been so badly mined that ground subsidence has been going on
for years. Other examples are numerous. Mining of coastal aquifers has led to
salt water intrusion into the aquifers such as: the Morphou (Güzel Yurt) area of Cyprus
and the Kino Bay area of northwestern Mexico.
COST OF EXPLOITATION
Where ground water is mined, the increased pumping cost resulting from mining increases
the pumping cost of water forever, and should be capitalized and compared with the
benefits obtained from the use of the water.
The cost of ground-water production from tubewells in the North Rohri project area in
Pakistan was estimated as Rs 1.52 per acre per acre-foot of water for five feet of
additional lift caused by mining the aquifer.
We have used the downloadable spreadsheet (English
units or Metric units) on this web site
for calculating the additional cost. We have made several assumptions. They
1. We adjusted the aquifer transmissivity to give five feet of drawdown.
2. The aquifer storage coefficient is 10 percent.
3. The Well operating factor is 80 percent.
4. The cost of electric power is 0.10 USD per kilowatt hour.
5. Overall Pump Efficiency is 70 percent.
6. Overall Motor efficiency is 70 percent.
7. One acre foot is produced over an eight month period.
The spreadsheet calculates the cost of producing the one acre foot of water as
$0.14/af. At an exchange rate of Rs 9 = 1.00 USD the cost of the water is Rs 1.26.
The actual cost estimated by other workers is Rs 1.52. The agreement is very good
given the uncertainties in input data and the exchange rate when pumping costs were
When capital is limited, we must choose a design criterion that achieves minimum
annual costs. Annual costs consist of amortized capital cost plus average annual
operating and maintenance costs. We believe minimum present value is a more suitable
measure because it takes account of the timing of operation and maintenance costs.
Thus, no average need be taken, and it also follows most closely the pattern required for
cash-flow estimation generally used for cost benefit analysis.
In a study of the Lower Indus Basin, it was shown that tubewell capital costs are a
linear function of depth and discharge, but operation and maintenance costs are quadratic.
Using the present value criterion, either optimum discharge or depth can be
determined. The appropriate discount factor is applied to the operation and
maintenance cost equation. The two equations are added and the differential of cost
with respect to discharge or depth is taken. The equations are solved for the
minimum cost by setting the first differential to zero.
This analysis shows water planners that the optimum discharge for a well is as large as
is feasible. Optimum depth, however, depends on discount
rates and aquifer transmissivity.
For low capacity wells, the marginal increases in overall cost as a function of
discount rates and transmissivity is small. But, as well capacities and overall
demand increase, the influence of aquifer transmissivity on overall cost increases
substantially. Enter into the spreadsheet your own data and compare it to your
actual cost experience.
The expected production capacity of wells will determine the suitability of expensive
ground-water exploration projects.
There can be little justification other than social and political policy for expensive
ground-water exploration projects for low-capacity village wells except in areas of
On the other hand, more expensive ground-water exploration projects can be easily
justified for high capacity wells for municipal, industrial and agricultural
Fortunately, there are different ground-water exploration methods to suit the range of
For example, AGW scientists have used fracture
trace analysis to successfully locate low capacity wells in West Africa very rapidly
and very inexpensively. In fact, in Burkina Faso the very first well we located
using fracture-trace analysis turned out to be very close to a well that another agency of
the government had located using more time-consuming and much more expensive earth
resistivity geophysical methods.
In Mali, prior to the beginning of the Mali Livestock Project, AGW scientists used very
inexpensive, hydrobiological ground-water-exploration methods. Hydrobiological
methods enabled us to pick well sites in the Sahel about as fast as we could drive through
For larger capacity wells for municipal and industrial purposes, AGW scientists have
used a combination of more costly hydrogeological investigation and Thermonic geophysical
When we evaluate the technical alternatives of ground-water-development programs we
must consider, among other things, the water requirements of the project and the number
and distribution of wells in relationship to hydraulic properties of aquifers, the
construction, operation and maintenance costs and the cost of capital.
ECAFE, 1971, Water Resources Journal, ST/ECAFE/SE.C/90