Challenges in African Hydrology and Water Resources (Proceedings of the Harare Symposium, July 1984). 1AHS Publ. no. 144.

Performance of sanitary completion measures of wells and boreholes used for rural water supplies in Malawi

W. J, LEWIS Department of Lands, Valuation and Water, Private Bag 311, Lilongwe 3, Malawi P. J, CHILTON* Hydrogeology Unit, British Geological Survey, Nallingford, Oxfordshire 0X10 8BB, UK

ABSTRACT A prime objective of the International Drinking Water Supply and Sanitation Decade is the worldwide provision of safe water supplies. If decade targets are to be met in Malawi, it is anticipated that, by the year 1990, there could be approximately 25 000 boreholes and wells serving about 75% of the rural population. In order to determine just how safe improved groundwater supplies are, a microbiological testing programme was conducted in two project areas during both the dry and wet seasons. The results show that boreholes typically deliver water with less than 10 faecal coliforms and less than 20 faecal streptococci per 100 ml and protected dug wells less than 20 faecal coliforms and less than 50 faecal streptococci per 100 ml. The authors consider that careful siting and good completion are the most approp­riate defences against faecal pollution of shallow ground­water supplies in Malawi.

Performances de mesures prises en exécution des règlements sanitaires concernant les puits et les forages utilisés pour 1'approvisionnement en eau de zones rurales du Malawi RESUME L'un des objectifs principaux de la Décennie Internationale pour l'Eau et l'Hygiène est l'approvision­nement en eau potable pour tous. Au Malawi, on prévoit qu'en 1990, si les objectifs de décennie sont atteints, il y aura environ 25 000 puits ou forages desservant environ 75% de la population rurale. Afin de déterminer dans quelle mesure la potabilité des eaux souterraines utilisée pour l'approvisionnement a été améliorée, un programme d'analyses microbiologiques a été réalisé dans deux zones de projets en saison sèche et en saison des pluies. Les résultats de ces recherches montrent que l'eau des forages contient moins de 10 conformes fécaux et moins de 20 streptocoques fécaux par 100 ml et que celle des puits protégés contient moins de 20 conformes fécaux et moins de 50 streptocoques fécaux par 100 ml. Les auteurs estiment que le choix du site et une bonne

^Present address: Department of Lands, Valuation and Water, Private Bag 311, Lilongwe 3, Malawi.

235

236 W.J.Lewis S P.J.Chilton

execution sont les meilleures défenses contre la pollution fécale des réserves d'eau peu profondes du Malawi.

INTRODUCTION

The 1980's have been declared the International Drinking Water Supply and Sanitation Decade. The Government of Malawi has adopted the aim of access to safe water for all its people and its Department of Lands, Valuation and Water is making substantial efforts to provide protected water supplies to the majority of the population living in rural areas. These efforts mainly involve three types of protected (untreated) water supply: gravity-fed piped water supplies, wells with handpumps and boreholes with handpumps. The former are drawn from catchments in mountainous areas with the in­takes located within protected forestry reserves1 free from habitation. However, it is anticipated that if rural water supply targets are to be met, there is a need for intensive and accelerated use of shallow groundwater resources. This paper assesses the ability of wells and boreholes to provide a safe water supply without treatment.

WATER QUALITY GOALS

In addition to making a supply of adequate quantity available for consumers, the purpose of providing a protected supply is to provide water which is safe to drink. A safe water is one that is free from disease-producing organisms (pathogenic bacteria and viruses) and highly toxic substances, although groundwater may contain unacceptable concentrations of dissolved minerals, e.g. fluorides, nitrates, sulphates. It is generally accepted, however, that the microbiological quality of drinking water is of fundamental importance and should never be compromised in favour of an aesthetically acceptable water. The International Standards for Drinking Water (WHO, 1971) are being revised and redrafted under the title of WHO Guidelines for Drinking Water Quality and in a recent paper Gorchev & Ozolins (1982) suggest that WHO will recommend a guideline of zero faecal coliforms and less than 10 total coliforms per 100 ml for unpiped supplies.

POLLUTANT PATHWAYS

The natural microbiological quality of groundwater, in areas of continuous soil cover, is generally very good, since soil is a very effective purification medium, having the ability to remove faecal micro-organisms and breakdown many chemical compounds. The soil cover and unsaturated zone are, therefore, the most important line of defence against pollution of underlying aquifers.

The severity of any contamination of a groundwater source, during its operating life, is closely related to its design and construction. Inadequate sanitary protection measures during construction allow pollutants to bypass the natural soil protection normally given to the aquifer. Ham (1971) suggests that failure to provide adequate

Performance of sanitary completion measures 237

sanitary protection to a well or borehole may be due to a number of reasons, among which are:

(a) ignorance on the part of the constructor of the geology of the area,

(b) the extra cost involved, and, (c) because the well is buried, "out of sight out of mind". There are two principal sources of potential faecal pollution of

rural groundwater supplies. These are aquifer pollution from excreta-disposal units such as pit latrines and direct well pol­lution from animal excreta and washing of clothes and cooking utensils. The factors governing the risk of groundwater pollution from on-site sanitation have been reviewed elsewhere (Lewis et al., 1982). The two key factors, determining the microbiological pol­lution risk are the thickness and nature of the soil overlying the aquifer (Fig.l). The risk of faecal groundwater pollution from pit latrines is minimal when the thickness of relatively fine uncon­solidated strata between the base of the latrine and the highest elevation of the groundwater table is greater than 2 m.

POROUS UNCONSOLIDATED (soils/sediments )

11111 M high risk unless covered by a minimum 2m of unsaturated I i 111II fine or medium-grained soils/sediments below latrine base

insufficient known to predicl risk with confidence

— D m increasing pollution risk

FIG.l Classification of soils and rocks in an array of relative microbiological pollution risk (after Lewis et al., 1982) .

HYDROGEOLOGICAL CONDITIONS

The two water-supply projects, from which the majority of samples in this testing programme have been taken, are located on the plateau of central Malawi. The plateau is underlain by ancient metaraorphic rocks, mainly gneisses, of the Pre-Cambrian Basement Complex. Prolonged weathering under tropical conditions has produced a shallow, extensive aquifer, which is of great importance as a source of potable water. The weathering profile may be up to 40 m thick and is characterized by a gradual transition upwards from fresh, unweathered gneiss through broken, hydrated, partially decomposed materials to a completely weathered, dominantly clayey surface layer. This relatively impermeable (10~2 to 10""% day-1) clayey cover produces

238 W.J.Lewis & P.J.Chilton

semi-confined aquifer conditions; the groundwater rises to a static level up to several metres above the depth at which it was struck during drilling. The properties of the weathered zone aquifer are described in more detail in an associated paper (Chilton & Smith-Carington, 1984).

In relation to microbiological pollution hazards, the most important characteristic of the weathered profile is the thick (8-15 m) clayey surface layer, which forms a highly effective barrier. In its natural state this aquifer is a low risk hydro-geological environment in terms of pollution vulnerability (residual soils in Fig.l). In addition, currently used separation distances between latrines and groundwater sources in the project areas are 100 m or more. The low permeability of the clay soil means that travel times are more than sufficient to eliminate faecal organisms. For these reasons, pit latrines are not, at present, considered a significant cause of faecal groundwater contamination in the plateau areas of Malawi. Evidence of bacteriological contamination of groundwater sources by faecal organisms is likely, therefore, to indicate localized pollution in the immediate vicinity of the groundwater source, and the effectiveness of sanitary protection measures must be examined.

Malawi has distinct wet and dry seasons. The wet season begins in November-December and continues intermittently until April, with the heaviest rainfall in January-February. The long dry season lasts from April until November. The plateau areas of central Malawi have an average annual rainfall of 800-1000 mm.

The current programme of rural groundwater supply projects in Malawi uses both boreholes and dug wells drawing from the weathered zone aquifer. The general philosophy has been that both have some advantages, as well as disadvantages, and could be used together in projects (Chilton et al., 1982). In the plateau areas there is a subdued topography of broad interfluves and shallow valleys, known locally as "dambo". Dug wells are restricted to the valleys and dambo edges where dry season water levels are within 3-5 m of the ground surface. Boreholes are constructed where groundwater levels are anticipated to be deeper than about 6 m. Both borehole and dug well sites are chosen by the villagers who will use them, under the general guidance of the hydrogeologist managing the project. The potential pollution hazards are explained to ensure, as far as possible, that these are avoided. An important outcome of this community involvement in the siting is that boreholes tend to be within, or very close to, the villages and dug wells tend to be more distant, often down in the dambo near to traditional sources.

GROUNDWATER SOURCE CONSTRUCTION TECHNIQUES

Protected dug wells (Figs.2 and 3(a))

Several variations of dug well design have been employed in Malawi; the two most relèvent to the present paper are shown in Fig.2. In both cases the shaft is first excavated by the villagers to the water level and then deepened under the supervision of project staff using a small motor driven de-watering pump, whose suction-lift

Perform

ance of

sanitary com

pletion m

easures 239

240 W.J.Lewis S P.J.Chilton

FIG.3 (a) Dug well completion; (b) borehole completion.

limits overall well depth to about 6.5 m. In this way the well is continued 2-3 m below the dry season water level to provide drought reliability. A significant sanitary design feature of some wells is that they are backfilled with excavated material to ground level (Fig.2).

Performance of sanitary completion measures 241

The surface works are similar in both cases (Fig.3(a)). The top slab is cemented onto the brick or stone wall and the cement facing of the wall extended over the brick apron surrounding the well, to provide a seal against pollution by waste water around the well itself. However, in the case of the brick lined well, the irregular annular space behind the bricks may not be effectively sealed by a proper cement grout. The waste water is conducted up to 10 m further away and downslope of the well by a brick and concrete drain, adjacent to which a washing slab is constructed. The waste water is then channelled into soakaway pits, which have proved ineffective because of the low-permeability clayey soils. The resultant standing water attracts animals and it is proposed to eliminate this problem by using the waste water to irrigate small vegetable gardens.

Protected shallow boreholes (Fig.3(b))

Borehole construction follows conventional design practices. Drilling is by cable tool percussion rigs, generally at 200 mm diameter. Depths range from 10 to 40 m, the average in the Llvulezi project is 24.5 m and in the Dowa West project 27.5 m. Groundwater levels are generally at 5-15 m depth, rarely as little as 3 m or as much as 20 m. The boreholes are completed with 110 mm, class 10, PVC pipe, slotted from the depth at which water was struck to the bottom of the borehole. The 50 mm annulus is filled with a gravel pack of 1-2 mm beach sand to well above the slotted section and the remaining annular space (perhaps 3-8 m) is filled with clay to 1 m below ground level. The remaining metre, which is often considerably oversize, is carefully filled with a concrete grout when the plinth is constructed for the handpump pedestal. The metre high pedestal of the new Malawi pump is set directly into the concrete plinth and the pumphead is bolted to it with a rubber gasket so that no waste water can be spilled actually at the wellhead itself. The above ground completion is identical to that of the dug wells (Fig.3(a)).

The total cost of a borehole complete with handpump serving about 250 people in these projects is approximately K100 per metre depth, i.e. about K2000 for a 20 m borehole. In comparison the cost of dug wells serving 125 people is K650 to K900, for the much smaller depth range of 4.0-6.5 m. The expenditure directly attributable to sanitary protection measures such as the extension pipe, plinth or well surround, apron and drain is about K150 to K200 in each case. This represents 20-25% of the total cost of dug wells and 5-10% of the total cost of boreholes.

METHODS OF MICROBIOLOGICAL EXAMINATION

For accurate bacteriological analysis, water samples should be pro­cessed as promptly as possible. Conditions can change after a number of hours of sample storage, so that samples may no longer be rep­resentative of conditions at the sampling site. The bacterial indicator organisms, like faecal coliforms, are particularly prone to "die-off" during storage. Standard Methods for the Examination of Water and Wastewater (American Public Health Assoc., American Water Works Assoc, and Water Pollution Control Federation, 1976),

242 W.J.Lewis S P.J.Chilton

advise against delays longer than 6 h between sample collection and sample incubation. The long distances between the central laboratory in Lilongwe and the widely scattered sampling sites thus preclude laboratory examination of water samples and necessitate field testing procedures.

Faecal coliform (FC) and faecal streptococci (FS) were enumerated, using the membrane filtration technique with Millipore field testing kits and incubators. Membrane enriched MF-C broth (Millipore) and incubation at 44.5°C ± 0.5°C for 24 h was used for the determination of faecal coliforms. Faecal streptococci determinations were made using the maltose azide membrane medium KF Streptococcus Agar (Difco), followed by incubation at 35°C + 0.5°C for 48 h. Testing was con­fined to the above mentioned organisms since it is generally recognized that these are more reliable indicators of faecal pollution than the coliform group as a whole.

RESULTS

Water samples were collected and tested from 36 dug wells in the Livulezi project area and 32 dug wells in the Dowa West project area during the month of October 1983 (end of dry season). These wells were retested in December 1983-January 1984 after the onset of the rainy season. Some 90 project boreholes from both areas were similarly tested during the two seasons. For comparitive purposes the results for some other sources routinely monitored during the period December 1982-December 1983 are also given. All these results are summarized in Table 1.

It is evident from Table 1 that the new project boreholes supply water of superior microbiological quality to all other sources. At least 80% of the boreholes tested during both seasons achieved the 0 FC/100 ml water quality goal; less than 4% of the boreholes were found to have in excess of 20 FC/100 ml. Next in order of superi­ority, at least in the dry season, were the backfilled dug wells, with some 81% achieving 0 FC/100 ml and only 5% exceeding 20 FC/100 ml. Only 63% of the dug wells tested which had not been backfilled achieved the zero standard, however, none of these wells exceeded 10 FC/100 ml. Although comparable in quality to project boreholes in the dry season the dug wells proved inferior in the wet season with only 50% of backfilled and 19% of non-backfilled dug wells recording 0 FC/100 ml. Even older boreholes of uncertain construction proved superior to dug wells in the wet season (Table 1).

The microbiological quality difference between boreholes and dug wells is more striking if one compares the data for faecal streptococci organisms (Table 1 and Fig.4). Only 59% of backfilled and 41% of the non-backfilled dug wells in the wet season achieved a standard of less than 20 FC/100 ml compared to 85% of boreholes. Faecal streptococci appear to be more predominant than faecal coliform organisms irrespective of season or water source type.

Faecal coliform organisms were found to be absent from only 5% of samples gathered from unprotected shallow wells, rivers and streams irrespective of season (Table 1). More than 40% of these unprotected sources had in excess of 500 FC/100 ml (Fig.5).

Perform

ance of

sanitary com

pletion m

easures 243

Si

•a

c !s 3

•a

cu o 3 o

o

i

eu -u m

&

o

i 3 o

•H

fH

fO

t>

o

En

O

q 0) 3 ty ai

3l

-U

•H

0)

CD

'a •H

o

D

O

O

O

+

J CM

CU

»H

•w

M <B

O

<u M

H

"a «

•H

M O

o

03 W

O

R

CD

O

nj

m

EH

HJ Q

) 01

i-H

+J

CD

1-H

cq

"O

p; q

En

m

., ai b

i c; m

fs eu

-C

! +

J

« "

H

î~H

S

O

O

>-s ~~\ o

&< -C

!

-u

-H

S

w

CD

O

U

3 O

co

*

?

O

O

in

A

o

o 1

-H

O

(Ni

O

o

CM

M

O

M

O

O

M 1

<-s LT|

Q

u~i I

M

CM

O

c\i [

•~s M

O

M [

>-i

O

•a eu

•u

01 CD

•u

6 « v—

CD

Q

, En

•u

eu

0 M

3 0 c/}

O

co

O

CO

O

O

00 "5f

O

O

O

O

10 CTi

O

CD

CO

Ci

IX

Ci

CM

c->

r-H

•H

O

co C

l

co in

m

H

C» CO

CO

N

I CO

O

O

O

O

"1

O

O

OO

C

i !

O

O

O

K

M

CM

^3" M"

Oi

l

CM LO

"*

O

•* I

Cl

co I

ON

CQ

(

^

•* [

in

°1

Ci

O

CM

c. &

c. &

Qfe

c, S: Q

&

q&

vo

CO

""̂

'a CD

M

•"H

;̂

MH

A

! 0 m

•C

l

01 |~H

i-H

eu

&

tn

3 Q

M<

CO

^

CM

c-i v~

'a eu

«

•H

M 01

M

M CU

&

tn

3 C

|

M

CO

Vw

M"

CM

O)

co

X

. s

-

01 eu

<-H

O

^ eu SH

0 rC

|

-u

CJ CU

•c

-i 0 in

Û

H

in

-qi

CM

C

M

x-

01 CU

•-H

.

0 X

! CD

M

0 -Q

'a M

O

CO

CM

-̂~

01 f-H

M

eu

à

"a eu

+J 0 eu

+J 0 CM

q

û

o, en

O

M

i-H

s

_

01 S

m

eu

M

+J 01

•a q

ta 01

eu

s> •H

«

CD tn q

us

CU

+J

q

••H

s o

o

>-f

\ +J •H

& 0]

CD

O

u 3 O

CO

O

co

tx C

i

CM

M1

Cl

Bi

"1 O

CM

Ci

O

CO

O

co CO

O

N

co

>-H

co co

i-H

C0

CO

CM

Q

•—s

co

CM

'a eu

i-S

M

;̂

HH

^ 0 •o

G 01

M

M CD

S

&) 3 d

>-H

co

CM

CM

&

•—s

CM

co

o

<"H

co

M

Ci

—̂

CM

co

'a eu

q

•H

M 0)

M

>-H

eu

&

Cn

3 C

,

Cl

>-H

CO

M

&

~̂

M

co

f-H

CM

M

[X

C,

-̂

M<

CM

01 eu

M

0 .q

CD

iH

0

-CI

+J 0 CD

0 M

On

CO

CM

C0

uo

Ss

~̂

ix M

>

season.

•u

eu S:

II

& • •*

q

0 01

eu

01

En

in

"a

H

c,

244 W.J.Lewis S P.J.Chilton

(a)

*- 50 i- 100 CD

O. 2S0 m 500 Li_ ,

500+

(b)

10 20 se

100 200 see

500+

.V. IVVVVVV IVW IV IVWVWV

Frequency (%) 40 60

* *

40 60

Frequency (%) 40 60

00 28 4S 60 y0 108

m <20 FS per 100 ml V . >20 FS per 100 ml * Cumulative frequency {%)

FIG.4 Microbial results for: (a) project boreholes in the wet season; (b) non-backfilled wells in the wet season.

(a) Frequency (%) 40 60

E

o

s_

H

C_>

(bl

e 1 8 2 0 5 0

1 8 0 2 8 8 5 8 8

500+

0

\

| § < *

m * m * iv * IVW * I V W W IVW I V W W W V W W

1 1 1

10 2 0

00 20

* *

V . V W W V [ 1 1

40 60

Frequency (% 4Q 6 0

_ 0 E 10 g 2 0 ^ 5 0 Si 100 °- 286 <_> s e e ^ 500+

I * I * I * IV * I W * I W W * I W W W I V W W W V W V W V W W

180

80 20 48 68 80 180

« <20 FC per 100 ml V . >20 FC per 100 ml * Cumulative frequency (%)

FIG.5 Microbial results from: (a) unprotected wells in the dry season; (b) rivers and streams in the dry season.

Performance of sanitary completion measures 245

DISCUSSION

During the dry season in the plateau regions of Malawi, a large proportion of properly constructed boreholes and wells meet the stringent WHO recommendation of 0 FC/100 ml.

The water quality from boreholes is better than from wells, whether backfilled or not. The water quality from backfilled wells is generally better than from non-backfilled wells particularly in the wet season; lack of a cement grout behind the lining of the latter may account for this difference in quality. Backfilled wells although providing water of better quality suffer the disadvantages of not allowing deepening in drought conditions and of having no means of extracting water with a bucket in the event of pump failure.

The inferior water quality of wells may be due to a number of reasons other than deficiencies in sanitary protection:

(a) Shallower groundwater tables (less than 2 m depth) with seasonal fluctuations which bring them near to ground level, in the areas where wells are typically sited.

(b) Presence of macro-fissures in the clay dambo soils, planar voids between peds and tubular channels formed by plants and animals living in the soil, allowing short circuiting of micro­biological contaminants to the groundwater table, especially with the first heavy rains.

(c) Greater potential for pollution since dambo are used for grazing and watering of livestock all the year round.

(d) Poor siting since dug wells are sometimes sited very close to traditional water sources which are always open and invariably grossly polluted.

Many of the groundwater sources tested were found to contain faecal streptococci even in the absence of faecal coliform. Certain investigations (Smith & Crabb, 1961; Geldreich et al., 1962) have shown that animals harbour in their gastro-intestinal tract a higher number of faecal streptococci than faecal coliform; the reverse is true in man. FC/FS ratios have often been recommended in stream pollution studies, as an index of the origin of contamination (e.g. Millipore, 1976). FC/FS ratios greater than 4.0 indicate pollution derived from human wastes whilst FC/FS of 0.7 or less indicate pollution derived from livestock or poultry. Geldreich et al. , (1968) have also found that faecal streptococci often persist longer than faecal coliforms, though this is not always true. For instance, McFeters et al., (1974) found that Streptococcus bovis and Streptococcus eguinis die-off considerably faster than faecal coliforms and other species of faecal streptococci. These are the dominant streptococcal species in some animal faeces, however, they never occur in human faeces. Thus, for meaningful ratios, samples should be tested within 24 h of the pollution event, because of different rates of bacterial die-off. For this reason it cannot be stated with any degree of certainty whether the predominance of faecal streptococci found in boreholes and wells is a result of faecal contamination by animals, since there is no way of determin­ing how long the organisms took to reach the groundwater, and how long they have been there. Field observations have shown that animal excreta is commonly found in the near vicinity of water sources and it is likely that animals are the main culprits.

246 W.J.Lewis & P.J.Chilton

However, further research including microbiological typing is required before this assumption can be proved. In any case, faecal streptococci appear to be the more sensitive indicator organism for the detection of faecal pollution of groundwater supplies.

Even though a considerable number of wells and some boreholes fail to meet the stringent WHO guideline value for small untreated water supplies of 0 FC/100 ml, a water with less than 20 or for that matter less than 50 FC/100 ml represents a considerable improvement in quality compared to water from unprotected wells, rivers and streams which normally contain in excess of 500 FC/100 ml.

The sanitary completion methods currently used in Malawi are proving effective. Good sanitary completion of boreholes and wells cannot guarantee total absence of contamination, this can only be achieved by chlorination. By the year 1990, if decade targets are to be met, it is anticipated that there could be approximately 25 000 wells and boreholes scattered throughout Malawi. The cost and logistics of supplying chemicals to treat these water sources would be formidable, especially if one considers that many earth roads are impassable during the rains.

It is considered that for the present, and in the forseeable future, careful siting and good sanitary completion will be the most appropriate defence against faecal contamination of shallow ground­water supplies in Malawi. Much attention should be paid to pre­venting contaminants reaching the groundwater table by short circuit­ing the natural protection afforded by the overlying soils, as is given to extracting the water in the first instance. In Malawi, boreholes and wells with the sanitary completion measures described in this paper will typically contain less than 10 and less than 20 FC/100 ml and less than 20 and less than 50 FS/100 ml respec­tively.

ACKNOWLEDGEMENTS This paper is published with the permission of the Malawi Government and the Director of the British Geological Survey. The authors wish to thank Dr S.S.D.Foster for suggesting the idea of this paper and for stimulating discussions during his field visits to Malawi. The field sampling and analyses were carried out by Mr P.G.S.Nkonjera, Mr R.H.Chimonja, Mr D.M.Kafulu, Mr A.F.Mwale, and Mr M.F.Chavula.

REFERENCES

American Public Health Assoc., American Water Works Assoc, and Water Pollution Control Federation (1976) Standard Methods for the Examination of Water and Wastewater. APH Assoc., AWW Assoc., and Water Pollut. Control Fed., Washington, USA.

Chilton, P.J. & Smith-Carington, A.K. (1984) Characteristics of the weathered basement awuifer in Malawi in relation to rural water supplies. In: Challenges In African Hydrology and Water Resources (Proc.-Harare Symp., July 1984), 57-72. IAHS Publ. no. 144.

Chilton, P.J., Grey, D.R.C. & Smith-Carington, A.K. (1982) Manual for Integrated Projects for Rural Groundwater Supplies. UNDP, Malawi.

Performance of sanitary completion measures 247

Geldreich, E.E., Bordner, R.H., Huff, C.B., Clark, H.F. &Kabler, P.W. (1962) Type distribution of coliform bacteria in the faeces of warm blooded animals. J. Wat. Pollut. Control Fed. 34, 295.

Geldreich, E.E., Best, L.C., Kenner, B.A. & Van Donsel, D.J. (1968) Tlie bacteriological aspects of stormwater pollution. J. Wat. Pollut. Control Fed. 40 (11), 1861-1872.

Gorchev, H.G. & Ozolins.G. (1982) WHO Guidelines for drinking water quality. In: Proc. International Water Supply Assoc. Congress, (Zurich). WHO.

Ham, H.H. (1971) Water wells and groundwater contamination. Bull. Assoc. Eng. Geol. 8 (1), 79-70.

Lewis, W.J., Foster, S.S.D. & Drasar, B.S. (1982) The risk of groundwater pollution by on-site sanitation in developing countries: a literature review. IRCWD - Report no. 01/82, International Reference Centre for Wastes Disposal, Dubendorf, Switzerland.

McFeters, G.A., Bisonnette, G.K., Jezeski, J.J., Thomson, C.A. & Stuart, D.G. (1974) Comparative survival of indicator bacteria and enteric pathogens in well water. Appl. Microbiol. 27 (5), 823-829.

Millipore (1976) Field Procedures in Water Microbiology. Cat. no. AB314, Millipore Corp. Bedford, Massachusetts, USA.

Smith, H.W. & Crabb, W.E. (1961) The faecal bacteria flora of animals and man: its development in the young. J. Pathol. Bacterid. 82, 53.

World Health Organisation (1971) International Standards for Drinking Water (3rd edn.), WHO, Geneva.