D.J. Allen

Aquifer properties of the Chalk of England

Aquifer properties data from 2100 pumping tests carried out in the Chalk aquifer have been collated as part of a joint British Geological Survey/Environment Agency project. The dataset is highly biased: most pumping tests have been undertaken in valley areas where the yield of the Chalk is highest. Transmissivity values from measured sites give the appearance of log-normality, but are not truly log-normal. The median of available data is 540 m2/d and the 25th and 75th percentiles 190 m2/d and 1500 m2/d respectively. Estimates of storage coefficient from unconfined tests have a median of 0.008 and from confined tests, 0.0006. The data indicate several trends and relationships in Chalk aquifer properties. Transmissivity is highest in the harder Chalk of Yorkshire and Lincolnshire (median 1800 m2/d). Throughout much of the Chalk aquifer a direct relation is observed between transmissivity and storage coefficient, reflecting the importance of fractures in governing both storage and transmissivity. Pumping tests undertaken in unconfined conditions give consistently higher measurements of transmissivity than in confined areas, probably as a result of increased dissolution enhancement of fractures in unconfined areas. At a catchment scale the data illustrate a relation between transmissivity and winter flowing streams.

Regional trends in matrix porosity and dry density of the Chalk of England

Abstract Laboratory measurements of porosity and dry density are presented for 2045 core samples from the Chalk of England. The data are subdivided on the basis of gross stratigraphy, i.e. Lower, Middle and Upper Chalk, and into four geographical areas: Northern England, East Anglia, Thames & Chilterns and Southern England. Statistical analysis of the data shows (i) that the porosity distributions for the Upper Chalk of the Southern and Thames & Chilterns regions are indistinguishable, (ii0 that the porosity distributions for the middle and Lower Chalk of the East Anglian region are indistinguishable, and (iii) that the porosity distributions for each of the gross stratigraphical units from all other regions are statistically discrete. Porosities range from 3.3% to 55.5%, with a mean porosity of 34.0%. Dry densities range from 1210 kg/m3 to 2510 kg/m3, with a mean dry density of 1790 kg/m3. In a given region there is a trend of increasing porosity from Lower to Middle to Upper Chalk. There are systematic variations in porosity between the regions. There is a trend of increasing porosity from the Northern England region to the Southern England region, to the Thames & Chilterns region, to East Anglia. No significant systematic variations in porosity-depth gradients were observed. Chalk porosity-depth gradients are typically high, of the order of -0.07 to -0.1 porosity per cent per metre.

Indirect detection of subsurface outflow from a rift valley lake

Abstract Naivasha, highest of the Kenya (Gregory) Rift Valley lakes, has no surface outlet. However, unlike other Rift lakes it has not become saline despite high potential evaporation rates, which indicates that there must be some subsurface drainage. The fate of this outflow has been the subject of speculation for many years, especially during the general decline in lake water level during the 1980s. Particularly to the south of the lake, there are few opportunities to obtain information from direct groundwater sampling. However, the stable isotopic composition of fumarole steam from late Quaternary volcanic centres in the area has been used to infer groundwater composition. Using a simple mixing model between Rift-flank groundwater and highly-evaporated lakewater, this has enabled subsurface water flow to be contoured by its lakewater content. By this method, outflow can still be detected some 30 km to the south of the lake. Stable isotope data also confirm that much of the steam used by the local Olkaria geothermal power station is derived from lakewater, though simple balance considerations show that steam use cannot alone be responsible for the fall in lake level observed during the 1980s.

Evidence for rapid groundwater flow and karst-type behaviour in the Chalk of southern England

Abstract With the growing importance of groundwater protection, there is increasing concern about the possibility of rapid groundwater flow in the Chalk of southern England and therefore in the frequency and distribution of ‘karstic’ features. Pumping test data, although useful in quantifying groundwater resources and regional flow, give little information on groundwater flow at a local scale. Evidence for rapid groundwater flow is gathered from other, less quantifiable methods. Nine different strands of evidence are drawn together: tracer tests; observations from Chalk caves; Chalk boreholes that pump sand; descriptions of adits; the nature of water-level fluctuations; the Chichester flood; the nature of the surface drainage; geomorphological features; and the presence of indicator bacteria in Chalk boreholes. Although the evidence does not prove the widespread existence of karstic features, it does suggest that rapid groundwater flow should be considered seriously throughout the Chalk. Rapid groundwater flow is generally more frequent close to Palaeogene cover and may also be associated with other forms of cover and valley bottoms.

Groundwater conceptual models: implications for evaluating diffuse pollution mitigation measures

The EU Water Framework Directive (WFD) identifies diffuse pollution as a long-term threat to water quality. Farming contributes significantly to this pollution. There is a clear need for mitigation measures and assessment of their efficacy. Accordingly, Demonstration Test Catchments (DTCs) have been established in England to test the effect of changes in agricultural practice on river water quality and ecology. However, the presence of groundwater in these hydrological systems implies a wide range of travel times for pollutants from source to receptor. Unless flow routes are better characterized, it will be difficult to gauge the success of control measures in the short term. Using 3D modelling and supplementary hydrochemical information, this study considers the hydrogeology of several sub-catchments in the Avon DTC, southern England. Data suggest that groundwater ages >25 years exist in parts of the catchments; clearly, observations like these must be used to judge the likely effectiveness of targeted control measures. The revealed hydrogeological complexity of the Avon catchment is unlikely to be unique, so the techniques described here should be applicable to other lowland river systems with moderate to high baseflow indices (>0.5). To support the WFD, groundwater conceptual models should inform the design of effective measures for diffuse pollution mitigation.

Geothermal exploration at Southampton in the UK: a case study of a low enthalpy resource

Two geothermal wells have been drilled in or near Southampton since 1979. One is at Marchwood, just outside Southampton, and the other is in the centre of the city; the wells are 1.85 km apart. They proved a geothermal reservoir in the upper 25 to 40 m of the Triassic Sherwood Sandstone at a depth of about 1700 m. The reasons for siting the wells in Southampton are discussed and the nature of the reservoir described. The wells have been extensively tested. The Marchwood well yielded 30 ls−1 for a pressure reduction of 3.7 MNm−2 after a test of 33 days, while the Western Esplanade well gave 20 ls−1 for about 3.0 MNm−2. Both yielded brine with a salinity of over 100 g l−1 at a well-head temperature of between 70 and 74°C. The transmissivity of the reservoir is 6 m2/d (3.5 D.m) and the storage coefficient 4 × 10−5. Computer modelling of changes in reservoir pressure suggests that near the wells there is a region of relatively high permeability but the permeability declines at distances of more than a few kilometres from the wells. This could take one of several forms including a bounded reservoir or a narrow wedge-shaped reservoir. The thermal yield from either well at an abstraction rate of 20 ls−1 would be about 3 MW.