Groundwater for Sustainable Development 1 (2015) 129134

Contents lists available at ScienceDirect

Groundwater for Sustainable Development

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

Functionality and water quality of Elephant pumps: Implications for

sustainable drinking water supplies in rural Malawi
Rochelle Holm a,n, Robert Stroud b,1, Golden Msilimba a, Simeon Gwayi a
a
b

Centre of Excellence in Water and Sanitation, Mzuzu University, P/B 201, Mzuzu 2, Malawi
Pump Aid, Development House, 56-64 Leonard Street, London EC2A 4LT, United Kingdom

art ic l e i nf o

a b s t r a c t

Article history:
Received 8 October 2015
Received in revised form
26 January 2016
Accepted 6 February 2016
Available online 10 February 2016

In 2013, our team performed a program evaluation of the sustainability and water quality of 338 Elephant
Pumps in Malawi. The team tested the water quality, evaluated the pump functionality and investigated
the community's satisfaction. The water in most samples (68%) conformed to the Malawi Government
drinking water standard for Escherichia coli. Likewise, the nitrate concentrations were within both the
Malawi Government and the WHO established standards. The functionality of the Elephant Pumps was
78% (producing water), which is comparable to the functionality rate of hand pumps in Sub-Saharan
Africa, which is posited as 64%. Pumps that are working well tend to have the lowest (signicantly) E. coli
contamination levels. The majority of the households use the water for common household activities;
however, water was also used for irrigation and commercial purposes in a limited number of cases.
Ninety one percent of the respondents reported that they were very satised with the pump design.
Although most Elephant Pumps produce water that conforms to the standards by the Malawi Government and with above average functionality, no perfect pump design exists. The performance of Elephant
Pumps can still be improved through better training of area mechanics and community pump caretakers,
and the availability of improved networks for spare parts.
& 2016 Elsevier B.V. All rights reserved.

Keywords:
Elephant pump
Functionality
Malawi
Water quality

1. Introduction
An estimated 663 million people are still without access to
improved sources of drinking water (UNICEF and World Health
Organization, 2015). In 2000, the United Nations proposed a Millennium Development Goal with respect to water quality of
halving the proportion of the population without sustainable
access to safe drinking water and basic sanitation by 2015 (United
Nations, 2015). Fortunately, Malawi has met the Millennium Development Goals for improved drinking water sources (UNICEF
and World Health Organization, 2015). However, diarrheal diseases, often associated with unsafe drinking water and poor sanitation, remain a leading cause of mortality in children under ve
in Malawi with over 3000 children dying per year (World Health
Organization). The study reported herein is an extension and
comparison of other research projects focused on rural Malawian
groundwater quality (British Geological Survey; Taylor et al., 2012;
n

Corresponding author.
E-mail addresses: rochelle@rochelleholm.com (R. Holm),
robtstroud_uk@yahoo.co.uk (R. Stroud), msilimba@yahoo.co.uk (G. Msilimba),
smgwayi@gmail.com (S. Gwayi).
1
Current Address: 1 Deepwell Drive, Camberley, Surrey GU15 2HF, United
Kingdom.
http://dx.doi.org/10.1016/j.gsd.2016.02.001
2352-801X/& 2016 Elsevier B.V. All rights reserved.

Pritchard et al., 2007; Pritchard et al., 2008; Wanda et al., 2011;

Kanyerere et al., 2012).
To address the gap in knowledge for provision of a safe rural
water supply, this study evaluated the functionality and water
quality parameters (total coliform, Escherichia coli, nitrate, total
hardness, total alkalinity, pH, free chlorine and total chlorine) in
shallow wells with Elephant Pumps (Fig. 1) installed by Pump Aid
in Malawi. This study has three goals: 1) to survey 338 Elephant
Pumps, including the oldest pumps, to identify possible sources of
contamination and to develop solutions to ensure the pumps are
providing safe drinking water, 2) to examine pump functionality
rates and downtime along with user satisfaction, and 3) to propose
recommendations for implementing and improving safe rural
water supply.

2. Materials and methods

For 6 weeks, a eld team covered ve districts and surveyed 338
Elephant Pumps (Fig. 2). During this time, the team tested the water
quality, evaluated the pump functionality, investigated community
satisfaction, and made any necessary repairs to the pumps.
The sampling strategy aimed to select a representative number
of pumps from 2843 pumps that were installed (at the time of the

130

R. Holm et al. / Groundwater for Sustainable Development 1 (2015) 129134

tables. In terms of inferential statistics, a one-sample t-test was

used to evaluate whether the mean for microbial parameters (E.
coli and total coliform counts) and chemical parameters (nitrate,
hardness and pH) in the water were signicantly different from
the Malawi Government (Malawi Bureau of Standards, 2005a,
2005b) and WHO (World Health Organization) standards. A Chisquare test showed the relationship between the visible pathways
for contamination and E. coli levels. The statistical analyzes were
conducted at the 95% condence interval.
Field data were collected from February to May 2013 during the
rainy season. Previous studies have shown that microbiological
contamination exceeding standards in Malawi was more prevalent
during the rainy season (Pritchard et al., 2007, 2008). This is also
supported by the earlier study from shallow wells and boreholes
in rural areas during the dry and rainy season by Taylor et al.
(2012), which found higher fecal contamination during the rainy
season. A study limitation was that for the study under Taylor et al.
(2012) data on well chlorination was not available.
Fig. 1. Elephant Pump, Malawi.

3. Results and discussion

study) by Pump Aid in Malawi; therefore, the results could be used
to draw conclusions on a countrywide basis. Three hundred thirty
eight pumps across the country were selected with respect to
physiographic area (113 in the highland/escarpment, 183 in the
plateau, and 42 in the wetland). The oldest pumps, installed in
2008 and 2009, were purposely chosen within each zone. Whether
a pump was functional or non-functional was unknown at the
time of study selection, although it was determined based on the
eld visit. A pump was classied as functional if water for sampling could be collected and drinking water was available to the
users on the day of the site visit.
After testing the survey tools, the data collection team used a
standardized questionnaire relying on both observations and focus
group discussions with the members of the water user committee
and community using the water.
Water quality analysis included total coliform, E. coli, nitrate, total
hardness, total alkalinity, pH, free chlorine and total chlorine. Of the
338 pumps studied, 285 were functional and available for water
quality sampling. In the sampling procedure, water was pumped from
the Elephant Pump to purge stagnant water prior to sample collection.
To capture a sample representative of drinking water conditions, the
sample was collected directly from the outlet of the pump into two
s
Whirl-Pak bags containing a sodium thiosulfate as a preservative
(Nasco, Fort Atkinson, Wisconsin). The samples were transported to a
eld lab for analysis the same day. The water samples were not ltered
prior to analysis. For the microbial analysis, new disposable Petri
plates, pipettes, and funnels were used once. The forceps were ame
sterilized, and the vacuum unit was wiped with alcohol and amed
before each new water sample was analyzed. Total coliforms and E.
s
coli were analyzed simultaneously using a Hach m-ColiBlue24 (Hach
Company, Loveland, Colorado) membrane ltration method in duplicate. Visible colonies were counted under 1015X magnication. For
chemical analyzes, the equipment was triple rinsed with sample water
prior to analysis. Nitrate was detected and quantied using a Hach
Nitrate Test Kit Model NI
11 (Hach Company, Loveland, Colorado).
The limit of detection is 1 mg/l, which is less than a tenth of the WHO
guideline, therefore providing adequate sensitivity (World Health Organization). The total hardness, total alkalinity, pH, free chlorine, and
total chlorine in the samples were monitored in duplicate with Hach,
Aquachek 5 in 1 Water Quality Test Strips (Hach Company, Loveland,
Colorado). The chemical assays were also performed in duplicate, and
the average results are reported.
The data were analyzed using statistical software, SPSS version
16, and Microsoft Excel. The data were summarized in terms of
means, medians, frequencies, and percentages, as represented in

3.1. Functionality
A majority of the pumps (69%) were reported as functioning
well (good yield and easily to be pumped by users), whereas 9%
were functioning badly (low water yield or difcult to pump) but
were still producing water, and 22% were not functioning. The
non-functioning wells were spread throughout the country and
were not localized to a geographic district. The percentage (78%) of
pumps producing water was compared with the hand pump
functionality rate in Sub-Saharan Africa, which was posited as 64%
based on data from 20 countries (Rural Water Supply Network).
The yield of most of the pumps (minimum 0.14 l/s; maximum
1.33 l/s) met the established government minimum of 0.25 l/sec
(Malawi Bureau of Standards, 2005a).
The functioning of hand pumps may be related to how long the
pump has been working combined with the ability of the community
to make repairs. Pumps reported as having low water yield or difcult
to pump water by users (9%) may be attributed to mechanical pump
malfunctions, initial site selection, or groundwater levels. Pump malfunction was primarily reported as the result of parts of the pump
wearing out, and lack of access to spare parts. In the current study, 69%
of the Elephant Pumps had broken down within their lifetime (ca. 67
years). The major causes of mechanical malfunction were due to a cut
rope (61.8%, n215) and rope slippage (18%, n63). Other lesser
causes noted included a broken pole and problems with the pipe. The
downtime of the pump was often only for a short period, and only 30%
of the pumps were reported to be out of order for over a week without
being repaired. Community members provided a number of reasons
for not repairing the pumps which were not functioning or functioning poorly. Data gathered using open-ended questions provided
the feedback that a lack of spare parts was the most frequent (73.3%)
reason for not repairing the pumps.
However, 87% of repairs were considered easy by the local
community members, including pump committee members, area
mechanics, and other local skilled personnel. The skills to repair
pumps are provided verbally through demonstrations by Pump
Aid during the installation. An analysis of the cost of the pump
repair revealed that 16% of the communities paid for the repairs,
and costs ranged from MK250MK6000 (0.38 d to 9.10 d/USD$0.54
to USD$12.95). The majority of the repairs cost between MK500
and MK2000 (0.76 d to 3.03 d/USD$1.08 to USD$4.31). Nonetheless, these pump functionality results compliment ndings in
the Democratic Republic of Congo that the availability of spare
parts is only a portion of hand pump sustainability, which requires

R. Holm et al. / Groundwater for Sustainable Development 1 (2015) 129134

131

Fig. 2. Map of Southern Africa/Malawi sample sites.

both a technical solution including repairs by local community

members as well as community motivation (Koestler et al., 2014.).
3.2. Water quality and national standards
The results for this study were compared to the standards for the
Malawi Government (50 cfu/100 ml) and WHO (0 cfu/100 ml)

(Table 1). For 13% of the wells, the WHO drinking water standard of
zero E. coli was met. The results indicated that more than half of the
samples (68%) had water equal to or better than the Malawi Government drinking water standard. These results are in contrast to an
earlier study by Taylor et al. (2012) which indicated only 12% of
Elephant Pumps had water equal to or better than the accepted
Malawi Government drinking water standard (n69). Remediation

132

R. Holm et al. / Groundwater for Sustainable Development 1 (2015) 129134

Table 1
Summary of the drinking water quality evaluated in Elephant Pumps in Malawi.
Parameter

E. coli
(colony forming units/100 ml)

Total Coliforms
(colony forming units/100 ml)

Nitrate
(mg/l NO3N)

Total Hardness
(mg/l as CaCO3)

Total Alkalinity
(mg/l as CaCO3)

pH

Minimum (n 285)
Mean (n 285)
Median (n 285)
Maximum (n 285)
Malawi Standarda
WHO Standardb

0
82
20
41000c
50
0

0
899
1000
410000c
50

0
1
1
7
10
11

25
161
120
425
800

0
99
80
250

6.2
6.5
6.2
8.4
6.0-9.5

No established value
a
b
c

Malawi Bureau of Standards (2005a, b).

World Health Organization
Result was too numerous to count, upper detection limit of method is reported.

(such as improving the surface seal, shock chlorine treatment of the

well and/or promotion of household point-of-use treatment) is recommended when the E. coli level is above the Malawi standard.
None of the Elephant Pumps in this study had been recently disinfected using chlorine, which shows a possible gap in the rural safe
water supply.
The nitrate guideline values present no risk of methemoglobinemia to infants less than 6 months who are not totally breast-fed.
The nitrate concentrations for all of the samples were signicantly
equal to or better than both the Malawi Government (10 mg/l NO3
N) and WHO established standards (11 mg/l NO3N) (minimum
value of 0 mg/l NO3N and a maximum of 7 mg/l NO3N). The
nitrate levels indicate that the Elephant pumps are not affected by
runoff from fertilizer or animals.
Typically, hard water is associated with the need for excess
soap for cleaning; however, excessive hardness may affect taste.
All of the water samples conformed to the Malawi Government's
hardness limit of 800 mg/l CaCO3. The total alkalinity in samples
ranged from 0 to 250 mg/l as CaCO3. Water samples from pumps
had a mean pH of 6.5, with a minimum of 6.2 and a maximum of
8.4. All of the pH values fall within the acceptable established
standard for the Malawi Government. The WHO does not have
established standards in terms for pH levels.
There are no established standards for free or total chlorine
levels in drinking water by the WHO or the Malawi Government.
The levels found in this study indicate that the water sources had
not been recently chlorinated to a level that resulted in residual
chlorine in them. Typically, shock chlorination of the wells is
performed by the government extension works in Malawi. These
results indicate the Ministry of Health district well inventory
chlorination program is not being conducted regularly. As a result,
it is important to consider residual chlorine with regard to overall
comparative results on water quality and not only the recent history for residual chlorine levels.
Kanyerere et al. (2012) surveyed six protected shallow wells in
northern Malawi, inclusive of Elephant Pumps and other designs,
and found that none had concentrations of any major ions of
concern when compared to WHO or Malawi Government drinking
water standards. Elephant Pumps require further study of water
quality data specic to human health issues of major ions and

trace metals.
Although national water quality standards exist, they are not in
line with the WHO standards and may not be sufcient to ensure
that improved water supplies are safe for drinking. Kayser et al.
(2015) suggested that changes in water quality governance in
Malawi are required, including coordination and data sharing
between ministries that deal with drinking water services; monitoring and enforcement of water quality laws; and sufcient
technical capacity to improve administrative and technical management of water services at the local level.
3.3. Factors affecting water quality
No statistically signicant relationship was found between
cracking on the pump casing and apron and E. coli levels. The
ndings show a greater proportion of water samples from pumps
that had no visible cracks in the cement on top or around the
pump superstructure cover had lower levels of E. coli than pumps
with cracks; 66% vs. 71%, respectively, had less than 50 cfu/100 ml
(i.e., 5% fewer). However, statistically the relationship between the
visibility of the cracks on the pump structures was independent of
the E. coli counts (X2(1, N 277) 0.198, p 0.657). The Chi-square
test showed the relationship between the visible pathways for
contamination and E. coli levels above or below the Malawi water
quality standard, and the results were not statistically signicant
at the 95% condence interval (i.e., at a 0.5 level of signicance, X2
(1, N 278) 0.534, p 0.465). In no cases were pumps observed to
be near mining or manufacturing operations.
3.4. Comparative pump type and water quality
Elephant Pumps are engineered with a tightened rope running
through a rope guide at the bottom of the well; the washers lift a
section of water with each pump stroke. In Malawi, other piston
pump types in shallow wells include the Malda and India Mark 5.
When combining the wet season data from Taylor et al. (2012)
with data from this study, Elephant Pumps provided water of
better quality (57%) compared to other piston pump types in
shallow wells of India Mark 5 (35%) and Malda pumps (41%) (Table 2). However, when comparing the Elephant Pump and

Table 2
Comparative wet season shallow well E. coli levels by pump type.
Elephant Pumps (this
study n 285)
% equal to or better than an E. coli 68%
level of 50 cfu/100 ml
cfu=colony forming unit(s).

Elephant Pumps (Taylor et al., 2012);

n 69 this study n 285 for total n 354)

India Mark 5 Pumps (Taylor

et al., 2012); n 26)

Malda Pumps (Taylor

et al., 2012); n153)

57%

35%

41%

R. Holm et al. / Groundwater for Sustainable Development 1 (2015) 129134

conventional piston pumps to shallow wells throughout rural

areas, none of the pump designs was able to provide safe water at
all times in a year.
3.5. Water uses
The designed maximum number of users for the Elephant
Pump is 24 households (120 people). For this study, the reported
average number of pump users was 22 households, which indicated that 71% of Elephant Pumps serve less households than the
design maximum. However, in 36% of communities, it was reported that some people were unable to use the pump, which was
primarily attributed to old age or disability. However, whether this
group had other water sources (protected or unprotected) or if
other community members gathered water for people unable to
use the pump, meaning that these people do use water from the
pumps but do not collect it themselves, requires further study. In
addition, how pump designs used by people with disabilities, including the Elephant Pump, with less or more mechanical difculty requires further study.
Community members primarily use water from the Elephant
Pump for common household activities. However, a few households use water for irrigation and commercial purposes, such as
making bricks and livestock. Ninety one percent of the respondents reported that they were very satised with the design
of the pump, most of whom highlighted the ease of use and repair.
When asked if they had any suggestions or recommendations, the
majority wanted spare parts and access to training for handling
the pumps. Therefore, improved information regarding where to
purchase spare parts and training of community members on repairing pumps should be implemented to improve safe rural water
supply. Regarding the perception of water quality, 90% of the respondents reported that the taste and color of the water was
acceptable.

133

This study represents a systematic approach for the examination of the functionality and water quality of Elephant Pumps in
Malawi. This study is unique and interesting based on the lack of
nationwide groundwater research and provides an extensive
shallow well study inclusive of physiographic areas including
highland/escarpment, plateau, and wetland. Critical actions that
are required to improve the sustainability of Elephant Pumps due
to mechanical malfunctions include making commercial spare
parts available at the regional level, including ropes, washers, and
rope guides, as well as strengthening the training of area mechanics and community pump caretakers. To ensure that the
pumps provide rural safe drinking water, one solution that could
be undertaken by government workers is regular disinfection
using chlorine to improve the microbiological quality after installation and at regular intervals, inclusive of each Elephant, India
Mark 5, and Malda pump designs. However, determining who pays
for regular disinfection of the shallow wells, whether Pump Aid,
the government, or the community, is an interesting study itself.
The government should also consider whether Elephant Pumps
should be an option as part of the planning and preventive health
programs in the future. Finally, national water quality standards
should be updated to the levels of WHO drinking water standards,
especially for microbiological parameters. Although no perfect
mechanical pump design exists for providing a continuously safe
rural water supply, this examination of their functionality and
water quality revealed that Elephant Pumps might be part of the
solution in Malawi.

Conicts of interest
This investigation received funding from Pump Aid Malawi.

Acknowledgments
3.6. Water point management and user satisfaction
Seventy eight percent of Elephant Pumps had a community
Water User Committee; the fact that the remaining 22% of the
pumps did not have a committee was attributed to the community
being uninformed about committee formation or the pump being
installed at a household who was responsible for managing it. Fifty
eight percent of the Water User Committees were collecting funds
from the user group. The monthly fee ranged from MK100 to
MK500 (0.15 d to 0.76 d/USD$0.21 to USD$1.08) per household.
This money is reportedly spent on grease, ropes, repairs, and the
well. The average amount of money spent by the Water User
Committee is MK1956 (2.97 d/USD$4.23). This indicates that the
repairs by the user groups are being successfully completed.

4. Conclusions
Elephant Pumps provide rural water at the community level in
Malawi, as evidenced by this investigation. In addition, most Elephant Pumps produce water that conforms to Malawi Government's water quality standard and with above average pump
functionality. No clear, environmental or anthropogenic, sources of
contamination were found that affect the water quality. Additionally, Elephant Pumps provide water of better quality compared to other approved piston pump types for shallow wells in
Malawi (Malda and India Mark 5 pumps). The majority of the
households used water for common household activities; however, water was also used for irrigation and commercial purposes
in limited cases. Ninety one percent of the respondents reported
that they were very satised with the pump design.

This research was rst presented at the 15th WaterNet/

WARFSA/GWP-SA Symposium, Lilongwe, Malawi, October 2931,
2014. The authors would like to thank the team of researchers who
covered many miles across Malawi during the 3 months of data
collection and Sue Cavill, Sally Sutton, Richard Carter, Huw Taylor
and James Ebdon for their support in the project design and nal
report review.

References
British Geological Survey. Groundwater Quality: Malawi. http://www.bgs.ac.uk/
downloads/directDownload.cfm?id 1283&noexcltrue&t Country%
20Sheets%3A%20Malawi (accessed 2.10.15).
Kanyerere, T., Levy, J., Xu, Y., Saka, J., 2012. Assessment of microbial contamination
of groundwater in upper Limphasa River catchment, located in a rural area of
northern Malawi. Water SA 38, 581596. http://dx.doi.org/10.4314/wsa.
v38i4.14.
Kayser, G.L., Amjad, U., Dalcanale, F., Bartram, J., Bentley, M., 2015. Drinking water
quality governance: a comparative case study of Brazil, Ecuador, and Malawi.
Environ. Sci. Policy 48, 186195. http://dx.doi.org/10.1016/j.envsci.2014.12.019.
Koestler, A.G., Kahorha, J., Biteete, L., 2014. Supply chain analysis of handpumps and
spare parts in Eastern Democratic Republic of Congo, UNICEF Contract
43147804, July 2014, A-Aqua AS, UNICEF, Oppegaard, Norway.
Malawi Bureau of Standards. Malawi standard; borehole and shallow well water
quality specication. Malawi Standards Board, MS 733:2005, 2005a.
Malawi Bureau of Standards. Malawi standard; drinking water specication.
Malawi Standards Board, MS 214:2005, 2005b.
Pritchard, M., Mkandawire, T., ONeill, J.G., 2007. Biological, chemical and physical
drinking water quality from shallow wells in Malawi: case study of Blantyre,
Chiradzulu and Mulanje. Phys. Chem. Earth 32, 11671177. http://dx.doi.org/
10.1016/j.pce.2007.07.013.
Pritchard, M., Mkandawire, T., ONeill, J.G., 2008. Assessment of groundwater
quality in shallow wells within the southern districts of Malawi. Phys. Chem.
Earth 33, 812823. http://dx.doi.org/10.1016/j.pce.2008.06.036.

134

R. Holm et al. / Groundwater for Sustainable Development 1 (2015) 129134

Rural Water Supply Network. Handpump data, selected countries in sub-Saharan

Africa. http://www.rural-water-supply.net/_ressources/documents/default/
203.pdf (accessed 2.10.15).
Taylor, H., Ebdon, J., Phillips, R., Chavula, G., Kapudzama, O., 2012. Assessment of
Drinking Water Quality For Lowcost Water Options In Rural Malawi Final
Report.
United Nations. Millennium development goal 7: ensure environmental sustainability. http://www.un.org/millenniumgoals/environ.shtml (accessed 2.10.15).
Wanda, E., Monjerezi, M., Mwatseteza, J.F., Kazembe, L.N., 2011. Hydro-geochemical
appraisal of groundwater quality from weathered basement aquifers in
Northern Malawi. Phys. Chem. Earth 36, 11971207. http://dx.doi.org/10.1016/j.

pce.2011.07.061.
World Health Organization. Child Mortality by Cause 20002013. http://www.
who.int/healthinfo/global_burden_disease/estimates_child_cod_2013/en/ (accessed 2.10.15).
World Health Organization. Guidelines for drinking-water quality, fourth edition.
http://www.who.int/water_sanitation_health/publications/2011/dwq_guide
lines/en/ (accessed 2.10.15).
World Health Organization. Progress on sanitation and drinking water 2015 update and MDG assessment. http://www.unicef.org/publications/index_82419.
html (accessed 2.10.15).