Characterising the vertical variations in hydraulic conductivity within the Chalk aquifer

Journal of Hydrology (2006) 330, 53– 62

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/jhydrol

Characterising the vertical variations in hydraulic conductivity within the Chalk aquifer A. Williams a b

a,*

, J. Bloomfield a, K. Griffiths a, A. Butler

b

British Geological Survey, Maclean Building, Crowmarsh Gifford, Wallingford, Oxfordshire, OX10 8BB, UK Department of Civil and Environmental Engineering, Imperial College, London SW7 2BU, UK

Accepted 12 April 2006

KEYWORDS

Summary Various field methods have been used to examine and quantify the vertical variations in aquifer properties within the Chalk aquifer at a LOCAR site in Berkshire, UK. The site contains three 100 m open boreholes and three sets of two nested piezometers within an area of about 100 m2. There is also an 86 m deep abstraction borehole about 40 m from the site. The techniques that have been used at the site include: geophysical logging, borehole imaging, packer testing, dilution testing and pumping tests. The packer test results show that the permeability of the aquifer varies by three orders of magnitude over the 70 m of tested material with a strongly non-linear decrease with depth below ground level. Comparison with the borehole images show that some of the highly permeable zones appear to be associated with obvious fractures. However, large fractures can be seen in zones which have much lower permeability while some highly permeable zones appear to be associated with poorly developed fractures. Single borehole dilution tests have shown that there are differences in flow velocity depth profiles over a few tens of meters across the site. These are inferred to be because the different boreholes, although of similar drilled depth and very close proximity, intersect slightly different parts of the fracture network and hence the groundwater flow system. In particular, a flowing feature at the base of one borehole is not intersected by the second, which is drilled from a slightly higher elevation. A dilution test carried out whilst the aquifer was being pumped shows that different fractures become active when the aquifer is stressed. This has implications for the interpretation of flow logs performed under pumping. c 2006 A. Williams. Published by Elsevier B.V. All rights reserved.

Chalk; Hydraulic conductivity; Tracer test; LOCAR



Hydrogeological context

* Corresponding author. Tel.: +44 1491 692294. E-mail address: [email protected] (A. Williams).



The Chalk is the most important aquifer in the UK (Downing et al., 1993) providing over half of all groundwater in the UK and locally in the south east of England up to 70% of all water in public supply. The requirements of the Water

0022-1694/$ - see front matter c 2006 A. Williams. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhydrol.2006.04.036

54 Framework Directive combined with a growing demand for water and other factors such as the increased frequency of hot dry summers in the last 10–15 years have led to pressure on Chalk groundwater resources. There is a need for more efficient groundwater source design and management, and one of the main challenges facing hydrogeologists working in this area is to optimise the use of groundwater given these changing environmental and social pressures. However, groundwater flow in the Chalk is highly heterogeneous and borehole yields may vary by orders of magnitude over distances of less than 100 m reflecting the complexity of flow in the aquifer. This paper reports on a field investigation aimed at increasing our understanding of controls on flow heterogeneity in the Chalk in the vicinity of groundwater sources. A characteristic feature of the Chalk aquifer of England is that significant permeability is commonly developed in or near the zone of water table fluctuation (Lloyd, 1993; Allen et al., 1997) and that there is a non-linear decrease in permeability with depth (Owen and Robinson, 1978). The Chalk is a dual porosity aquifer (Price et al., 1993) consisting of a microporous matrix intersected by fractures. Due to the small pore-throat size of the matrix (typically about 0.1–1 lm, Price et al., 1976), it has a very low permeability (Allen et al., 1997) and does not contribute to regional groundwater flow. Therefore, it is features of the fracture porosity that principally control variations in permeability and it is variations in fracture characteristics with depth that influence the significant changes in vertical permeability. Price (1987) distinguished two types of fracture porosity in the Chalk; primary fracture porosity associated with unmodified fractures and secondary fracture porosity associated with fractures inferred to be enlarged by carbonate dissolution. Fracture apertures in secondary fractures are larger and hence more transmissive. Bloomfield (1996) has suggested that other factors may contribute to the development of secondary fracture porosity, but whatever their genesis, it is commonly assumed that it is the secondary fractures that have the most influence on rapid flow in the Chalk and that control the distribution of permeability within the aquifer (Downing et al., 1993). A number of field investigations in the 1970s and 1980s used a combination of TV logs, flow logs, and packer test to investigate vertical variations in Chalk permeability (Tate et al., 1970; Headworth, 1972; Foster and Milton, 1974; Foster and Robertson, 1977; Price et al., 1977, 1992; Owen and Robinson, 1978; Connorton and Reed, 1978). Allen et al. (1997) summarized the main observations of these studies as follows: • Permeability measured in boreholes using packer tests and other field scale methods is usually at least an order of magnitude higher than matrix permeability measured on core plug samples illustrating the contribution of flow in the primary fracture component of the fracture network. • Most of the saturated thickness of the Chalk in a given borehole has low permeability with only a few intervals providing a significant contribution to the observed transmissivity. These may correspond to intervals with fractures.

A. Williams et al. • The major flow horizons are concentrated near the top of the saturated Chalk in the zone of water table fluctuation with little significant flow deeper than 50 m below the groundwater level. This is due to a decrease in fracture density and fracture aperture with depth because of the increasing overburden and a general reduction in groundwater circulation and hence potential for development of secondary fracture porosity. • Geological heterogeneities, hardgrounds, flints and lithological boundaries can locally increase the permeability of the Chalk. It is clear from the previous field studies that flow near individual boreholes is highly heterogeneous, but that there is great uncertainty in the relationship between the characteristics of a fracture observed in a borehole and the amount of flow which that fracture contributes to the borehole. In addition, few of the previous studies explicitly considered the influence that the borehole has on the observed flow. Using the LOCAR Programme field site at Trumpletts Farm, the aims of this study were to bring together a wide range of field investigation techniques to characterize vertical variations in the rock mass and hydraulic characteristics of the Chalk, to assess the relationships between fracture characteristics and the observed flow and head distributions, and to investigate the aquifer under natural and stressed conditions to see what this may reveal about the flowing features which dominate under these contrasting conditions. Following a description of the field site and the programme of work, the results of the packer testing and single borehole dilution tests are presented.

Description of site and field testing programme The testing described in this paper was performed at LOCAR site PL10 in the Pang Lambourne catchment (Wheater et al., 2006). The site, known as Trumpletts Farm, consists of six boreholes drilled in the vicinity of an Environment Agency abstraction borehole (confusingly known as Bottom Barn). Three of the boreholes are completed with pairs of piezometers whilst the other three were left open, apart from surface casing. Completion details are given in Table 1. The site is located on the side of a dry valley on the Chalk outcrop (Seaford Chalk formation at outcrop). It slopes gently from north to south, with an elevation difference of about 5 m between Borehole B and the abstraction borehole (Fig. 1). The water table at the site is about 20 m below ground level and has an annual fluctuation of about 7 m. The water table slopes gently in the same direction as the land surface but with a difference of only 0.5 m between Borehole B and the abstraction borehole. This gives a hydraulic gradient of about 0.001. All of the boreholes were drilled in the Chalk, with some soil in the upper few metres. Geophysical logs for Borehole A are shown in Fig. 2. The Chalk Rock can be easily identified at a depth of 82 m below datum (25.8 m aOD). A comprehensive field testing programme was planned for this site. The techniques used were packer testing, pump testing and tracer testing. Each of these techniques is described in detail below.

Characterising the vertical variations in hydraulic conductivity within the Chalk aquifer Table 1

55

Completion details of the boreholes at Trumpletts Farm

Borehole number

Drilled depth (m bGL)

Elevation (m AOD)

Max casing depth (m)

Piezometer depth (m)

Borehole diameter (mm)

PL10A PL10B PL10C

100 100 40.2

107.97 110.51 110.26

18 20 3

143 143 194

PL10D

39.9

108.05

3

PL10E PL10F

100 40.8

107.47 108.12

20 2.5

EA BH

86.0

105.75

25

Open hole Open hole 30.1–29.8 40.1–39.6 27.0–26.8 40.0–39.8 Open hole 30.0–29.8 40.0–39.8 Open hole

Packer tests Packers are expanding plugs that can be used to isolate temporarily a section of a borehole. The packer system used at Trumpletts Farm (Price and Williams, 1993) incorporates a pump, to abstract water from the isolated section, and a transducer to measure the pressure within the section. This system allows measurements to be made of the permeability of the isolated section and the head within the section. This latter measurement means that a profile can be determined of the natural head within the formation, undisturbed by vertical flow within the borehole. When the packers are inflated, isolating a section of the aquifer from the borehole, the water level in both the isolated interval, and the section of borehole above the interval may change from that which was measured in the open hole. The head below the isolated interval may change as well but it is not possible to monitor this with the present packer system. The water level in an open borehole represents a ‘weighted average’ of the head at the different depths penetrated by the hole. The ‘average’ is weighted by the permeability of the different contributing horizons. For example, if there

B C D E F A

EA Open borehole

N

Piezometer

0

Figure 1

Site map.

20 m

(C2) (C1) (D2) (D1) (F2) (F1)

194 143 194 760

are two contributing layers A and B of the same thickness with permeabilities of KA and KB, and heads of HA and HB, the open borehole head (HO) will be given by: HO ¼

K A HA þ K B HB KA þ KB

The head within the aquifer at any depth changes with time. What is measured with a packer system is a ‘relative head change’. Within a section this is recorded as the change in water level, which occurs in that section when it is isolated. If the interval has a lower head than that present in the borehole then a negative change is noted. The implication of a negative change is that when the packers are not present, water will flow from the borehole into the formation. The rate of this flow can be calculated, as the permeability of the isolated section will have been measured during the subsequent packer test (flow rate is proportional to the permeability and the head difference). Once the packers have inflated isolating the section to be tested, water is pumped out of the isolated interval at a constant rate. The head in the section is monitored using the transducer and pumping is continued until a steadystate drawdown is measured. This usually takes about 20 min. The pumping rate is then increased and the head monitored until a new steady-state is reached. The rate is then decreased and a further steady-state achieved. Ideally the permeability calculated for each of these tests will be similar enough for confidence to be placed in the test. After this the packers are deflated and lowered to a new interval. This whole process takes at least 2 h (depending on how many pumping rates are used and how long it takes to achieve steady-state) which means that it is difficult to carry out more that three tests in a day. As usual with fieldwork all did not progress as planned. It was intended to packer test boreholes A, B and E to provide a continuous measure of permeability from 30 to 100 m bgl which would be used to indicate how this parameter varies vertically and laterally within an approximate 100 · 100 · 100 m cube. The first problem encountered was that the drilled diameter of Borehole A, in the top 20 m below the casing, was larger than the packer system could cope with. Thus, it was not possible to test this important interval of the Chalk at this location with the existing packer system. Over three testing cycles, 13 3 m intervals were eventually tested in the lower part of Borehole A.

56

A. Williams et al. Well Name: Trumpletts Farm Bh A Location: 451315 175037 Datum at 107.8 aOD Geophysical logsmrun by BGS April and May 2005. Depth (mbd) 5

BGS Caliper (inch)

10 0

Natural Gamma (API Cs) 20 8

Induction COND (mS/m)

0

Fluid TEMP 26Apr05 -8 10.5 11.5 Fluid EC25 520 (25-Oct-04) 545

-10

-20

-30

-40

-50

-60

-70

-80

-90

-100

Figure 2

Geophysical logs for Borehole A.

The next problem was encountered when Borehole E was tested. It had not been possible to perform a single borehole dilution test in this borehole as the conductivity probe would not drop down the hole. However, a subsequent calliper log showed no obstructions or any significant changes in diameter from the casing to the total drilled depth, and no difficulty was encountered in lowering the logging sondes. Packer testing of the top two sections of the borehole proceeded normally, but the packer string could not be lowered to the third interval. Following this, it also proved difficult to lift the string out of the borehole. Thus, this borehole was not available for any further testing. Following the difficulties experienced in Borehole E, it was decided

not to try and lower the packer string past an obstruction encountered at about 65 m aOD in Borehole B, meaning that only 5 intervals were tested in this borehole. In summary, the plan to produce comprehensive data on the hydraulic conductivity variations at the site was thwarted by circumstance.

Single borehole dilution and pumping tests Several single borehole dilution tests have been performed at Boreholes A and B, and Borehole A was also tested whilst the EA abstraction hole was being pumped. A single borehole dilution test (SBDT) is a tracer test carried out in a

Characterising the vertical variations in hydraulic conductivity within the Chalk aquifer single borehole. The intention of the test is to give an indication of the groundwater velocity in the vicinity of the borehole. SBDTs (Freeze and Cherry, 1979; Gaspar, 1987) are used with some success in shallow boreholes in intergranular aquifers, but the technique has not been widely used in deep boreholes in UK aquifers. The technique involves introducing a uniform column of tracer into the open (or screened) length of a borehole and monitoring the dilution of the tracer with time. This makes it a tracer technique that is guaranteed of a positive result as it does not depend on anticipating the direction or speed of travel of the tracer. For the tests at Trumpletts Farm the tracer used was common salt (NaCl) and its concentration was monitored by measuring the conductivity of the water in the borehole. The salt was introduced in a large diameter hose, which was carefully removed to leave a column of saline water within the borehole (Ward et al., 1998). A conductivity probe was then lowered into the borehole and a series of conductivity logs run. The log called ‘Initial’ was run as soon as possible after the hose was removed from the borehole, but the logistics of removing the hose slowly enough to give reasonably uniform tracer injection meant a delay of around 20 min from when the hose removal begins to when the logging can

Table 2

begin. The logs are run at about 5 m/min which means that each logging run takes 15–20 min. The time recorded for other logging runs is the time they were started after the start of the ‘Initial’ run. A short (7 h) pumping test has been carried out with manual water level measurements taken at all of the boreholes and piezometers, and a further pumping test and radial flow tracer test has been carried out very recently. The results from these tests will not be reported here, as further testing, including packer testing at the top of Borehole A with larger packers, is still to be performed at the site.

Results Packer tests The packer test results from Trumpletts Farm are summarised in Table 2 and the hydraulic conductivity measurements from Borehole A are shown in Fig. 3. Also shown in Table 2 are the head measurements made during the packer testing. The values shown in Table 2 are the change in head within the isolated interval when the packers are sealed. A positive value means that the head in the interval is higher than that of the open hole, which implies that when the

Summary of packer testing results

Borehole

Borehole A

Date

March 2003

November 2003

March 2004

Borehole B

March 2004

Borehole E

March 2004

a

57

Interval Top (m AOD)

Bottom (m AOD)

64.97 59.47 53.97 48.97 37.47 32.47 29.47 23.97 19.97 14.97 66.47 64.97 56.47 63.44 57.44 46.44 44.44 41.44 35.94 29.94 27.44 83.99 81.24 76.32 72.19 69.19 83.54 80.64

62.07 56.57 51.07 46.07 34.57 29.57 26.57 21.07 17.07 12.07 63.06 61.56 53.06 60.03 54.03 43.03 41.03 38.03 32.53 26.53 24.03 80.58 77.83 72.91 68.78 65.78 80.13 77.23

Permeability (m/d)

Relative head change (m)a

13 41 1.5 5.3 0.14 0.08 3.0 0.04 0.09 0.14 3.2 27 0.72 0.86 0.48 0.57 6.4 7.0 0.57 5.0 0.50 180 >500 1.7 14 1.8 58 27

0.003 0.024 +0.080 +0.083 +0.073 +0.006 +0.004 +0.048 +0.011 +0.017 0.014 0.009 +0.084 0.003 0.038 +0.019 +0.054 +0.031 +0.005 0.018 0.02 0.027 0.075 +0.047 0.013 0.006

A positive value means that the head in the packered-off interval is higher than that in the open borehole. This means that water will flow into the borehole when the packers are not in place.

58

A. Williams et al. Packer Test Results At Borehole A 100 Mar-03 Nov-03 Mar-04 90

80

70

m aOD

60

50

40

30

20

10

0 0.01

0.1

1

10

100

K (m/day)

Figure 3

Packer test results.

packers are not present water would flow into the hole at this interval. The measurements taken suggest that in Borehole A water is flowing into the borehole at the bottom and out above 56 m aOD. A similar scenario seems to occur in Bore-

hole B, but with the flow into the borehole occurring at about 66 m aOD. However, it must be remembered that not all of the length of the open boreholes have been tested and so significant horizons may not be represented in these results.

Characterising the vertical variations in hydraulic conductivity within the Chalk aquifer The permeability of Borehole A (Fig. 3) can be seen to vary over three orders of magnitude with permeability reducing strongly non-linearly with depth. The results also show high variability within short depth intervals, particularly in the upper tested sections. This is in part due to the variability in the permeability of the Chalk and the length scale of the tested intervals (3 m). The packers are not emplaced at exactly the same depth every time and so a particularly transmissive feature may be tested on one occasion but not the next time the section is tested. There is a general trend of permeability reducing with depth, but consideration of the results in the 40–50 m aOD section suggests that the permeability is probably the result of the presence of discrete fractures rather than a more generally permeable rock mass. This is consistent with the general understanding of the hydraulic behaviour of the Chalk. For the reasons given above, there are no results available from the upper section of Borehole A. The upper section of Borehole B has been tested and shows the highest permeability value measured at this site, and the top of Borehole E also has high permeability. This general feature of very high permeability in the zone of water table fluctuation is also consistent with the accepted understanding of the Chalk aquifer. To try and define better the features that contribute to the permeability in the Chalk, the visual logs of the borehole were examined for the intervals tested. These logs were acquired using an Optical Televiewer Probe (Robertson Geologging) which provided a continuous, orientated 360 image of the borehole walls using an optical imaging system. Fig. 4 shows the scanned images of the borehole wall

Figure 4

59

for three of the packer test intervals in Borehole A. The first picture shows the image of the highly permeable zone at 58 m aOD and the section shows features that would be recognised as potentially significant hydraulic features. The second picture shows, in contrast the featureless Chalk that would be expected of a poorly permeable interval. However, the third image, which shows some apparently significant features, is of another poorly permeable section. This implies that not all of the visually identifiable features are hydraulically active. This phenomenon has been observed in sandstone aquifers (Price, 2003) and the usual explanation for this is that the fractures do not persist very far. It is also possible in the Chalk that the features identified visually are caused by the removal of small flints during the drilling of the borehole.

Single borehole dilution and pumping tests The results of the first single borehole dilution tests carried out in Boreholes A and B are shown in Fig. 5. It is apparent from Fig. 5 that Boreholes A and B behave very differently. The initial profile for Borehole A shows a significant dilution has occurred in the upper 10 m and in the bottom 5 m, before the log was run (assuming uniform injection of the tracer). There is then little change in the shape of the profiles between 60 and 30 m aOD in the next 46 min, and nearly all the tracer has left after 20 h. The differences between the initial and 46 min profiles suggest that there is flow into and then up the borehole from about 12 m aOD, which creates a freshening profile up the hole. Significant dilution also

Images of tested intervals with different K values.

60

A. Williams et al. Inferred flow directions

100

90

90

80

80

70

70

60

60

50

Inferred flow directions

Borehole B

100

m aOD

m aOD

Borehole A

50

40

40

30

30

20

20

10

10

0

0 0

1000

2000

3000

4000

0

5000

1000

Conductivity (µS/cm) Background

Initial

After 46 mins

2000

3000

4000

5000

Conductivity (µS/cm) Background

After 20 hrs

Figure 5

Initial

After 96 minutes

After 21 hrs

SBDT result for BHs A and B.

occurs between 58 and 75 m aOD, implying that there is flow across and out of the borehole. Flow in the borehole is caused by head variations within the aquifer. As measured during the packer tests there is a significant difference in head between the upper and lower parts of the aquifer which will cause water to flow into the borehole at the bottom and out at the top. The location and magnitude of these flows will depend on the exact location and permeability of the hydraulically active features intercepted by the borehole. There is also flow across the borehole due to the local hydraulic gradient. In conventional (intergranular medium) SBDTs it is these flows that dominate the profiles and allow an estimate to be made of local groundwater velocities. It is clear from these profiles that this is not the case in this part of the Chalk aquifer. Borehole B shows a different type of behaviour to that shown by Borehole A. It is notable that the initial profile is much more dilute than that of Borehole A over a significant length of the borehole. Also, a substantial amount of tracer still remains in the bottom part of the borehole after 20 h. This implies faster flowing groundwater in the top part and slower flow in the bottom. As the two boreholes are only 30 m apart this is surprising. However, examination of the logs from the very bottom of the boreholes (Fig. 6) shows what may be causing the difference in observed behaviour. There is obviously a significant inflow of water between 10 and 12 m aOD in Borehole A that is not present in Borehole B. This causes an upward flow in Borehole A that flushes the tracer towards the upper part of the hole (and then out to the aquifer). As this inflow is not present in Borehole B the tracer remains within the borehole until flushed out

by the regional groundwater flow. That the feature providing this flushing is not intercepted by Borehole B can be explained by the fact that, although drilled to the same depth below ground, Borehole A penetrates slightly further into the formation due to the difference in the elevation of the boreholes. Following these initial SBDTs, two further tests were carried out in Borehole A. These were done directly before and during a short pumping test carried out at the EA abstraction borehole (Fig. 1). The profiles obtained during these tests are shown in Fig. 7. The first thing to notice from these profiles is the similarity between the ‘before pumping’ test and the test discussed above. This is significant as the tests were carried out at different times of year, (the first in November and the second in March) when it might be supposed that the regional hydraulic conditions would be different. This profile again shows the feature at about 12 m aOD where water appears to flow into the borehole and a further feature at about 58 m aOD that limits the upward flow of the tracer. The profiles taken whilst the EA borehole (40 m away) was pumping at 6 Ml/d show a different behaviour, with the feature at 75 m aOD becoming dominant and no indication of any feature at 58 m aOD. However, it is clear that profiles should be run at shorter time intervals before any real comparisons can be made.

Discussion and Conclusions Packer tests measure in situ properties of the rock which are difficult to obtain by any other means and as such provide a valuable contribution to our knowledge of Chalk

Characterising the vertical variations in hydraulic conductivity within the Chalk aquifer Borehole A

18

16

16

m aOD

m aOD

20

18

14

14

12

12

10

10

8

8 0

1000

2000

3000

4000

5000

0

1000

Conductivity (µS/cm) Background

Initial

2000

3000

4000

5000

Conductivity (µS/cm)

After 46 mins

Background

After 20 hrs

Figure 6

Initial

After 96 mins

After 21 hrs

SBDT result for bottom of BHs A and B.

Borehole A Before pumping

Borehole A During pumping

Inferred flow directions

100

90

90

80

80

70

70

60

60

50

50

40

40

30

30

20

20

10

10

0

Inferred flow directions

100

m aOD

m aOD

Inferred flow directions

Borehole B

Inferred flow directions

20

61

0 0

1000

2000

3000

4000

5000

0

1000

Conductivity (µS/cm) Background

Initial

After 44 mins

Figure 7

2000

3000

4000

5000

Conductivity (µS/cm) After 113 mins

Background

Initial

After 47 mins

SBDT result for BH A before and during pumping.

After 95 mins

After 202 mins

62 behaviour. However, these tests require specialist equipment which is not readily available and are time consuming to carry out. As the work at Trumpletts Farm has shown the success of the technique is also dependant on the quality of the borehole. The repeatability of the tests needs to be further investigated, as usually only one set of tests would be performed. However, the comparison with borehole images has been instructive and shows that images can be misinterpreted if hydraulic data are not available. SBDTs are easy and quick to carry out, requiring minimal equipment. However, the interpretation of the test results from deep boreholes needs some more consideration. In general, only one borehole would be tested at a site and so the differences shown between the two boreholes of slightly different effective depths would not have been seen. The tests reported here also highlight the differences in flow paths, which may be dominant when the aquifer is stressed by pumping. The main conclusions can be drawn from these initial tests at this site are: • The Chalk aquifer is fairly heterogeneous at this site. • The more different types of tests that are used, the more information is obtained and greater certainty is achieved. • Testing the aquifer under stressed conditions does not necessarily reveal much about the flowing features which dominate during natural conditions. It is important to use field data appropriately and to consider whether different interpretations are required, for example for source evaluation and for understanding aquifer properties required for regional modelling work. Further work carried out at this site will build on these experiences and will hopefully allow a better picture to emerge of the effect the heterogeneity in the Chalk properties at this scale has on the overall behaviour of the aquifer.

Acknowledgements The work reported on in this paper was carried out as part of the NERC LOCAR Project NER/T/S/2001/00941 ‘Investigation of groundwater flow heterogeneity in Chalk aquifers using detailed borehole arrays and stochastic modelling techniques’. The field work could not have happened without the hard work of Jude Cobbing, Alex Gallagher and Magali Moreau of BGS, for which the authors are very grateful. This paper is published with the permission of the Executive Director of the British Geological Survey.

References Allen, DJ., Brewerton, LJ., Coleby, LM., Gibbs, BR., Lewis, MA., MacDonald, AM., Wagstaff, SJ., Williams, AT., 1997. The physical properties of major aquifers in England and Wales. British Geological Survey Technical Report WD/97/34, Nottingham, UK. pp. 312.

A. Williams et al. Bloomfield, J.P., 1996. Characterisation of hydrogeologically significant fracture distributions in the Chalk: an example from the Upper Chalk of southern England. Journal of Hydrology 184, 355–379. Connorton, B.J., Reed, N.R., 1978. A numerical model fro the prediction of long term well yield in an unconfined chalk aquifer. Quarterly Journal of Engineering Geology 11, 127–138. Downing, R.A., Price, M., Jones, G.P., 1993. The Hydrogeology of the Chalk of North-West Europe. Clarendon Press, Oxford, UK. Foster, S.S.D., Milton, V.A., 1974. The permeability and storage of an unconfined aquifer. Hydrological Sciences Bulletin 19, 485–500. Foster, S.S.D., Robertson, A.S., 1977. Evaluation of semi-confined Chalk aquifer in East Anglia. Proceedings of the Institution of Civil Engineers 63, 803–817. Freeze, R.A., Cherry, J.A., 1979. Groundwater. Prentice-Hall, Englewood Cliffs, NJ, pp. 604. Gaspar, E. (Ed.), 1987. Modern Trends in Tracer Hydrology, vol. II. CRC Press, Boca Raton, FL. Headworth, H.G., 1972. The analysis of natural groundwater level fluctuations in the Chalk of Hampshire. Journal of the Institute of Water Engineers and Scientists 26, 107–124. Lloyd, J.W., 1993. The United Kingdom. In: Downin, R.A., Price, M., Jones, G.P. (Eds.), The Hydrogeology of the Chalk of North-West Europe. Clarendon Press, Oxford, UK, pp. 220–249. Owen, M., Robinson, V.K., 1978. Characteristics and yield in fissured Chalk. Institution of Civil Engineers, Symposium on Thames Groundwater Scheme paper 2, 33–49. Price, M., 1987. Fluid flow in the Chalk of England in fluid flow in sedimentary basins and aquifers. In: Goff, J.C., Williams B.P.J. (Eds.). Special Publication of the Geological Society of London, vol. 34. pp. 141–156. Price, M., 2003. The origin and extent of some fissures in sandstone (Extended Abstract). In: Proceedings of the International Conference on Groundwater in Fractured Rocks. International Association of Hydrogeologists, Prague, pp. 87–88. Price, M., Williams, A.T., 1993. A pumped double-packer system for use in aquifer evaluation and groundwater sampling. Proceedings of the Institution of Civil Engineers Water, Maritime and Energy 101 (June), 85–92. Price, M., Bird, M.J., Foster, S.S.D., 1976. Chalk pore-size measurements and their significance. Water Services (October), 596–600. Price, M., Robertson, A.S., Foster, S.S.D., 1977. Chalk permeability – a study of vertical variations using water injection tests and borehole logging. Water Services 81, 603–610. Price, M., Atkinson, T.C., Barker, J.A., Wheeler, D., Monkhouse, R.A., 1992. A tracer study of the danger posed to a Chalk aquifer by contaminated highway run-off. Proceedings of the Institution of Civil Engineers Water, Maritime and Energy 96 (March), 9–18. Price, M., Downing, R.A., Edmunds, W.M., 1993. The Chalk as an aquifer. In: Downing, R.A., Price, M., Jones, GP. (Eds.), The Hydrogeology of the Chalk of North-West Europe. Clarendon Press, Oxford, UK, pp. 14–34. Tate, T.K., Robertson, A.S., Gray, D.A., 1970. The hydrogeological investigation of fissure flow by borehole logging techniques. Quarterly Journal of Engineering Geology 2, 195–215. Ward, R.S., Williams, A.T., Barker, J.A., Brewerton, L.J., Gale, I.N., 1998. Groundwater tracer tests: a review and guidelines for their use in British aquifers. R&D Technical Report W160, Environment Agency, UK. Wheater, H.S., Neal, C., Peach, D., 2006. Hydro-ecological functioning of the Pang and Lambourn catchments, UK: an introduction to the special issue. J. Hydrol., this issue, doi:10.1016/ j.jhydrol.2006.04.035.