Evaluation of aquifer characteristics of Dhaka city using pumped well data

Abstract

Dhaka WASA has divided the whole city into six zones to administer its operations of potable water supply and safe sewerage disposal. The construction and the development test data of 44 selected tubewells from these six zones have been analyzed to evaluate its aquifer and well specific properties. Parameters, such as, transmissivity and storativity are evaluated to characterize the aquifer properties while well efficiency and skin factor are evaluated to demonstrate its ~ell specific parameters. Another important parameter called specific capacity is also determined to represent the productivity of both the aquifer and the well itself. The development tests on the deep tubewells of DWASA are performed without installing piezometers. Thus, the pumping tests conducted by DWASA can be treated as a single well test. Determination of transmissivity and storativity are therefore constrained. with the utilization of the well-known methods based on piezometric readings. To utilize these single well tests under the confined and/or leaky aquifer system, there are 5 suitable methods selected from literature to overcome the lack of piezometer readings. The selected 5 methods are Jacob's straight line method, Theis's recovery method, Papadopulos-Cooper's method (curve fitting), Hurl' -Worthington's method and the Logan approximate method. Among these 5 methods, Jacob straight line method gives consistently higher transmissivity values than the other 4 methods. Jacob method is very sensitive to the non-linear well losses which have been found in appreciable magnitude for the most of the wells tested in Dhaka city. Papadopulos-Cooper method is valid for fully penetrated confined aquifer, mostly not the case existing for the configuration and aquifer system analysed here. Hurr- Worthington and recovery methods have been judged the most suitable methods for the aquifer system in Dhaka city. Out of these two, Hurr- Worthington method is selected to draw the isotransmissivity lines to represent its variations instantaneously in Dhaka city. It is seen from. the map (Fig. 5.1) that the south-western part and north-eastern corner of Dhaka city have higher transmissivity (700-1000 m2/d) values than the rest of the city. The majority of the city area representing transmissivity values ranging from 400 to 500 m2/d. The Hurr- Worthington method evaluates the transmissivity by using the corrections due to well losses

on its observed drawdowns. While the aquifer loss is dependent on the aquifer characteristics only, the well loss arises from the combined effect of screen entrance loss and frictional loss inside the pumped wells. The 'aquifer losses in Dhaka city are found to vary from 8.2 111 to 20.9 m and the total well losses from 2.1 m to 14.7 m. As the well losses are found appreciable in Dhaka city compared to its overall drawdowns, it dictate the use of an appropriate method which takes care of these losses in its transmissivity determination. In this respect, .Hurr- Worthington method may be a better option to be used for Dhaka city aquifer system. The Logan method, known as an approximation met~od uses the specific drawdown data for transmissivity determination is liable to produce some error under vertical leakage and transient aquifer responses. For the selected single well methods, the storativity values can only be provided by the Papadopulos-Cooper method and its values vary from 1.5x 10-6 to 9. 5x 10-5 . It is se,en from the iso-storativity map (Fig. 5.2) that the south-western part of Dhaka city have the hIghest storativity values than the rest of the city. The majority of the city area representing storativity values of about bCl 0-5. Both the specific capacity and the well efficiency are important to measure the performance of a well. ,In Dhaka city,. the specific capacity of the wells varies from 219 m2~day to 550 m2/day and well efficiency varies from 45 to 86 percent. Low efficiency as low as 70 percent is the indication of inadequate and poor well development: Again, skin effect is to account for the head losses in the vicinity of a well. Skin factor in Dhaka city is found to vary 0.32 to 8.06 .