Page 314 - Hydrogeology Principles and Practice

HYDC08 12/5/05 5:31 PM Page 297

Groundwater resources and environmental management 297

Fig. 8.11 Cross-section, drawn
perpendicular to the river, showing
the idealized conceptual model of
river–aquifer interaction for a confined
aquifer. Prior to pumping the initial
potentiometric surface (shown as line
RWL, the rest water level) is horizontal
and equal to the constant water level of
the stream. The drawdown due to
pumping is shown as line PWL, the
pumping water level.

Although the assumptions behind the solution of (that is, finite-width stream of shallow penetration
equations 8.3 and 8.5 are an over-simplification of adjoining an aquifer of limited lateral extent) is pre-
reality, analytical results can provide rough estimates sented by Butler et al. (2001). The solution shows
of the local impacts of abstraction on river flow and that the conventional assumption of a fully penetrat-
the timescales over which flow depletion occurs. ing stream can lead to significant over-estimation of
By neglecting river bed and river bank sediments stream depletion (>100%) in many practical situ-
and assuming full aquifer penetration, the impact of ations, depending on the value of the stream leakance
pumping on the stream flow is over-estimated and parameter and the distance from the pumping well
the time delay between abstraction starting and the to the stream. An important assumption underlying
impact of pumping on stream flow is underestim- this new solution is that the penetration of the stream
ated. A further assumption that prior to pumping the channel is negligible relative to aquifer thickness,
initial potentiometric surface is horizontal and equal although an approximate extension to the method
to the constant water level of the stream (Fig. 8.11) is, provides reasonable results for the range of relative
in fact, not a major limitation (Kirk & Herbert 2002), penetrations found in most natural systems (up to
but Wilson (1993, 1994) presented steady-state ana- 85%); Butler et al. (2001).
lytical solutions for two-dimensional, vertically integ- To assist in the practical application of analytical
rated models of induced infiltration from surface solutions, stream depletion caused by groundwater
water bodies for various combinations of aquifer abstraction can be readily calculated using dimen-
geometry in the presence of ambient aquifer flow. sionless type curves and tables. Jenkins (1968) pre-
Wallace et al. (1990) extended the approach of sented a number of worked examples, including
Jenkins (1968) to show that pumping impacts may computations of the rate of stream depletion for the
develop over several annual cycles. For cases where pumping and following non-pumping periods, the
the stream depletion impacts develop over long time- volume of water induced by pumping and the effects
scales due to either the distance between the bore- (both rate and volume of stream depletion) of any
hole and river, the type of aquifer properties or the selected pattern of intermittent pumping. An ex-
possible role of river deposits, the maximum impact ample calculation is given in Box 8.4.
in later years may exceed the maximum depletion in
the first year. Such delayed impacts are potentially an
important catchment management consideration in 8.3.2 Catchment resource modelling of
order to avoid future low river flows. river flow depletion
A new analytical solution for estimation of draw-
down and stream depletion under conditions that Local-scale impacts of groundwater abstraction
are more representative of those in natural systems on river flows can be investigated with the above