Research Report Sustainable WASH Systems Learning Partnership A WATER INFRASTRUCTURE AUDIT OF KITUI COUNTY Cliff Nyaga, University of Oxford January 2019 Research Report PHOTO CREDIT: CLIFF NYAGA/UNIVERSITY OF OXFORD
Research Report
Sustainable WASH Systems Learning Partnership
A WATER INFRASTRUCTURE AUDIT OF KITUI COUNTYCliff Nyaga, University of OxfordJanuary 2019
Research Report
PHO
TO C
RED
IT: C
LIFF
NYA
GA
/UN
IVER
SITY
OF
OX
FORD
2
Prepared by: Cliff Nyaga, University of Oxford
Reviewed by: Mike Thomas, Rural Focus; Eduardo Perez, Global Communities; Karl Linden, University of
Colorado Boulder (UCB); and Pranav Chintalapati, UCB.
Acknowledgements: The Kitui County Government would like to acknowledge the financial support received
from the United States Agency for International Development (USAID). Further, the Kitui County Government
appreciates its longstanding partnership with the University of Oxford and UNICEF Kenya through various
collaborating programs, including the DFID-funded REACH Program.
The leadership received from Emmanuel Kisangau, Kennedy Mutati, Philip Nzula, Augustus Ndingo, and Hope Sila
— all from the County Ministry for Water Agriculture and Livestock Development — throughout the audit
exercise is appreciated. The sub-county water officers were instrumental in logistics planning and in providing
liaison between the field audit teams, communities, and County Ministries for Agriculture, Water, and Livestock
Development and Administration and Coordination. A team of local enumerators led field data collection: Lucy
Mweti, Grace Muisyo, Abigael Kyenze, Patrick Mulwa, Lydia Mwikali, Muimi Kivoko, Philip Muthengi, Mary Sammy,
Ruth Mwende, Peter Musili, Annah Kavata, James Kimanzi, Purity Maingi, Felix Muthui, and Assumpta Mwikali.
The technical advice and guidance received from Professor Rob Hope of the University of Oxford and Dr. Andrew
Trevett of UNICEF Kenya throughout the planning, data collection, analysis, and preparation of this report is very
much appreciated.
Front cover: This Katanu Hand pump was developed in the late 1990s by the Government of Kenya and is the
main water source for Nzamba Village in Ikutha Ward, Kitui. Photo credit: Cliff Nyaga/University of Oxford.
About the Sustainable WASH Systems Learning Partnership: The Sustainable WASH Systems Learning
Partnership is a global United States Agency for International Development (USAID) cooperative agreement to
identify locally-driven solutions to the challenge of developing robust local systems capable of sustaining water,
sanitation, and hygiene (WASH) service delivery. This report is made possible by the generous support of the
American people through USAID under the terms of the Cooperative Agreement AID-OAA-A-16-00075. The
contents are the responsibility of the Sustainable WASH Systems Learning Partnership and do not necessarily
reflect the views of USAID or the United States Government. For more information, visit
www.globalwaters.org/SWS, or contact Elizabeth Jordan ([email protected]).
A Water Infrastructure Audit of Kitui County 3
Table of Contents
Acronyms............................................................................................................................................................................. 6
Foreword ............................................................................................................................................................................. 7
Executive Summary ........................................................................................................................................................... 8
Introduction ...................................................................................................................................................................... 13
Kitui County Water Audit ............................................................................................................................................ 17
Approach and Methodology .......................................................................................................................................... 19
Audit Results ..................................................................................................................................................................... 24
Opportunities for Sustainability of Rural Water Services ..................................................................................... 56
Conclusion ......................................................................................................................................................................... 62
Annex A: Kitui County Water Audit Methodology ................................................................................................ 63
List of Figures
Figure 1. Map of Kenya (left) and map of Kitui County showing the eight sub-counties (right) .................. 14
Figure 2. Sample report on the ONA dashboard providing a quick analysis of audit data ............................ 21
Figure 3. Photo menu of ONA web showing image data submitted by enumerators .................................... 22
Figure 4. Distribution of the 3,126 water sources located mapped through the water audits in 2011,
2016, and 2017 ......................................................................................................................................................... 24
Figure 5. Distribution of water sources mapped by type, number, and proportion of the total 3,126
sources ....................................................................................................................................................................... 25
Figure 6. Distribution of the 3,126 water sources mapped by sub-county ....................................................... 27
Figure 7. Kitui seasonal rainfall anomalies (left) and distribution of infrastructure (earth dams, hand
pumps, and rock catchments) developed over time (right) .......................................................................... 28
Figure 8. Reported annual reliability of surface and sub-surface dams ............................................................... 28
Figure 9. Functionality status of all water sources mapped on the day of audit............................................... 29
Figure 10. Functionality of all water sources mapped on the day of audit by source type ............................ 30
4
Figure 11. Functionality status of all water sources mapped and counted by sub-county ............................. 31
Figure 12. Kitui County investors by number of water sources developed ..................................................... 32
Figure 13. Leading investors in leading water infrastructure installed since 2000 ........................................... 33
Figure 14. Cumulative unadjusted investment (USD) in leading source types since 2000 ............................. 34
Figure 15. Cumulative unadjusted USD investment made in piped schemes from 2000 to 2017 ............... 35
Figure 16. Functionality of hand pumps across the different management options documented ................ 36
Figure 17. Leading cause of mechanical breakdown in non-operational hand pumps registered ................. 37
Figure 18. Dry periods for the 32 percent of hand pumps that reported unreliable supply ......................... 37
Figure 19. Water distribution pressure for piped schemes (left) and the primary energy source(s) in
pumped schemes (right) ........................................................................................................................................ 38
Figure 20. Operation status of piped schemes by the energy sources used for water pumping ................. 39
Figure 21. Reported broken features in partly-functional and non-functional piped schemes on the day of
the water audit ......................................................................................................................................................... 41
Figure 22. Reported cost of major repairs completed in 2 years and the volatility of repair costs ............ 42
Figure 23. Actors involved in fixing the most recent breakdown in piped schemes ....................................... 43
Figure 24. Downtime for all failures reported by piped schemes during the audit ......................................... 44
Figure 25. Common preventive maintenance tasks done in rural piped schemes ........................................... 44
Figure 26. Leading reasons for scheme close-down at certain times of the year ............................................ 45
Figure 27. Reliability against condition of piped scheme ........................................................................................ 46
Figure 28. Maximum one-way distance (kilometers) travelled by households to piped scheme sources
(left) and density of piped scheme investments per capita and per km2 (right) ....................................... 47
Figure 29. Variation of TDS (ppm) levels across the eight sub-counties of Kitui ............................................ 49
Figure 30. Domestic and productive water uses recorded in the audit ............................................................. 49
Figure 31. Reported key monthly expense items in piped schemes by value ................................................... 50
Figure 32. Reported collection efficiency range by piped schemes for billed revenues ................................. 51
A Water Infrastructure Audit of Kitui County 5
Figure 33. Documented management options in piped schemes ......................................................................... 51
Figure 34. Number of employees by the employing entity.................................................................................... 52
Figure 35. Proportion of piped scheme employees occupying observed roles ................................................ 53
Figure 36. Mean monthly salary reported by job/role in piped schemes ........................................................... 54
Figure 37. Mean monthly salary (USD) for piped scheme staff against education attained ........................... 54
Figure 38. Reported primary and secondary alternative sources of drinking water for piped scheme
users ........................................................................................................................................................................... 58
Figure 39. Functionality of piped schemes by sub-county ...................................................................................... 60
List of Tables
Table 1. Number of water sources mapped by type, and their distribution by sub-county ......................... 26
Table 2. Number of major breakdowns recorded in piped schemes in 2 years .............................................. 42
6
Acronyms
CBO Community-Based Organization
DFID Department for International Development
KIMWASCO Kiambere-Mwingi Water and Sanitation Company
KITWASCO Kitui Water and Sanitation Company
MAWL Kitui County Ministry of Agriculture, Water, and Livestock Development
O&M Operation and Maintenance
ONA A web and mobile app product for field data collection
PPP Public-Private Partnerships
SDG Sustainable Development Goal
SHG Self-Help Group
SWS Sustainable WASH Systems
TDS Total Dissolved Solids
UCB University of Colorado Boulder
USAID United States Agency for International Development
WASH Water, Sanitation, and Hygiene
WASREB Water Services Regulatory Board
WMC Water Management Committee
WSP Water Service Provider or Water Utility
7
Foreword
This county water audit has been conducted at a strategic point in time when we really want to
determine how much water is accessible to the people of Kitui. Under the leadership of Her
Excellency, The Governor Charity K. Ngilu, provision of clean and reliable water is the development
manifesto pillar number one for Kitui County Government. This is informed by the multiple benefits
that come with access to reliable and safe water: improved health for the population, improved
access to good nutrition, and creation of time for other livelihood activities. It is a common thing in
many rural areas of Kitui to see women and even children sitting in long queues waiting to fetch
water. Sometimes this water is not easily accessed, and the people are forced to dig deep and risky
sand wells to get water. It is the desire of our governor through the Water Department to make
this water available at the homestead level and remove the burden on women of carrying water with
their backs. Due to the expansiveness of Kitui County, water access is still low in terms of coverage
and quantity. Quite a significant population walk long distances in search of water and queue for long
hours waiting to draw water. This situation is unacceptable and forms our thrust in planning for
investment.
This report has produced a very useful inventory of both functional and non-functional rural water
facilities to facilitate evidence-based decision making. It shows the facilities to direct our investments
toward, and at the same time it underscores the need for inclusion of a kitty for maintenance in the
budgeting process. The report also identifies key actors and their contribution in the water and
sanitation sector, and this forms a useful database for strengthening coordination and hence
improving service delivery. As I have observed in just about a year in this office, for smooth
implementation of WASH programs, a proper information management system is paramount. Data
management is a key factor in giving us baselines, setting targets, and monitoring progress over time.
The county water audit provides a rich data set that will require frequent updating by all WASH
actors in the county. This will ultimately improve sectoral coordination and inform improved
planning and better policies and practices, which will accelerate Kitui County’s achievement of the
sustainable development goals to ensure universal water access by 2030. My office will strive to
establish this information system and always encourage relevant collaboration by all partners.
I want to take this opportunity to register my sincere appreciation to the organizations who have
given their best energy, time, expertise, and financial resources to make this work a success. The
support that we have received so far and continue to enjoy from the University of Oxford, UNICEF
Kenya, and Rural Focus through the USAID Sustainable WASH Systems Learning Partnership is
immense and highly appreciated. I will not forget to thank my team in the Water Department for
being available to support this process whenever they are needed.
Thank you all.
Emmanuel Kisangau,
County Minister for Agriculture, Water, and Livestock Development
8
Executive Summary
Kitui County is one of 47 counties in Kenya, located approximately 160 kilometers (km) east of
Nairobi City. The county has an estimated population of 1,221,000 people, with 14 percent located
in urban areas and 86 percent in rural parts of the county. The county established two water service
providers (WSPs), Kitui Water and Sanitation Company (KITWASCO) and Kiambere-Mwingi Water
and Sanitation Company (KIMWASCO). The WSPs are licensed and regulated by the Water
Services Regulatory Board (WASREB) to manage water and sanitation services in Kitui County.
According to WASREB, in 2017, KITWASCO and KIMWASCO served 32 percent of the county
population (approximately 360,000 people), in predominantly urban areas due to limited network
coverage in rural parts of the county. County-wide estimates indicate that 42 percent of the
combined rural and urban population have access to at least a basic water service. Performance of
the rural water sub-sector has not been comprehensively documented; hence, an information gap
exists on water coverage and quality of service for a large segment of the county population.
To fill this gap, the Sustainable WASH Systems Learning Partnership (SWS), funded by the United
States Agency for International Development (USAID), and the REACH Program, funded by the
Department for International Development (DFID), conceptualized and supported a water audit in
2016 and 2017. The water audit aimed to locate major rural water infrastructure and collect
information on installation and operational performance to inform county planning, investment,
institutional development, and dialogue on sustainability.
The University of Oxford developed a water audit instrument with the county government and
piloted it in Mwingi North Sub-County. The team recruited and trained local enumerators and
collected data using the Open Data Kit mobile platform in Mwingi North Sub-County in August
2016, and for Kitui’s remaining seven sub-counties from November to December 2017. The water
audit targeted major and equipped infrastructure, namely hand pumps and piped schemes, and non-
equipped sources that skirted the equipped infrastructure. The analysis in this report therefore
reflects major infrastructure, rather than a systematic survey, and it is possible that not every water
source in the county was located and mapped.
Results of the Water Audit
Cumulatively, the audits mapped 3,126 equipped and non-equipped water sources spread across
Kitui County. In terms of number of equipped sources, 687 hand pumps (22 percent) and 460 piped
schemes (15 percent) were audited. In terms of non-equipped sources, 655 sand dams (21 percent),
613 earth dams (20 percent), 268 shallow wells (9 percent), 89 rock catchments (3 percent), 28
springs (1 percent), and 52 incomplete boreholes (2 percent) were documented. On the day of the
audit, 60 percent of all sources mapped were fully functional and 40 percent were either partly
working (15 percent) or not working at all (25 percent). By number of sources developed, NGOs
are the leading investor (37 percent) in county water infrastructure, followed by national and county
governments (23 percent) and community-funded development (10 percent). For 11 percent of the
sources mapped, respondents credited funding for infrastructure development to bilateral and
multilateral donors. More bilateral and multilateral donors than reported here may have invested in
the Kitui County water sector, given that they constitute a primary grants source for NGOs. In
terms of value of investment, piped scheme sources account for more than half of the total
investment (unadjusted) made in Kitui County since 2000, followed by earth dam sources. For 19
A Water Infrastructure Audit of Kitui County 9
percent (86) of piped schemes where cost information was available, the total value of the capital
investment was $5.22 million (unadjusted).
Reliability of water sources is a key issue for Kitui County, stemming from a climate-related lack of
year-round availability of water at sources mapped. Climate analysis suggests that there is a high
inter-seasonal rainfall variability and extreme drought risk in some parts of the county. When paired
with infrastructure mapped, the analysis suggests that many surface water sources (i.e., rock
catchments, earth dams, and hand pumps) have been constructed in areas of sequential dry periods.
Results for Hand Pump Sources
A hand pump was defined as a well or borehole with hand-operated mechanical pumping equipment
installed for groundwater abstraction. Afridev is the dominant hand pump type in Kitui County (86
percent), followed by the India Mark II hand pump (7 percent). Two out of three hand pumps supply
water all year round, while the remainder, 32 percent, experience periods of decreased flows,
especially in the months of August, September, and October. Pump rods (32 percent), foot valves
(28 percent), and rising main pipes (31 percent) constituted the most broken parts in non-
operational hand pumps.
Results for Piped Scheme Sources
A piped scheme was defined as a water source (whether borehole, rock catchment, shallow well,
earth dam, or sand dam) that incorporates a water distribution network and/or includes pumps,
tanks, valves, meters, pipeline, kiosks, and standpipes, among other features. In total, 460 piped
schemes were audited, 454 of which were fully completed and six under construction. On the day of
the audit, 56 percent of the completed piped schemes were fully functional, 15 percent were partly
functional, and 28 percent were not working. The most broken features in the piped scheme
constituted the pipe network (17 percent), generators (13 percent), boreholes (12 percent), pumps
(10 percent), and storage tanks (10 percent). Twenty-five percent of the breakdowns experienced in
the 2 years of the audit were resolved within 7 days, while it took up to 30 days to resolve 46
percent of the breakdowns reported. On average, pipe network and vandalism-related failures took
a shorter time to fix (169 and 159 days, respectively) compared to pump failures, which took 352
days on average to fix. Kitui County Government is the most active entity involved in fixing
breakdowns, for 32 percent of the piped schemes, followed by local technicians, for 30 percent of
the schemes. Regarding reliability of service, the average duration of water supply for fully
operational piped schemes at the time of the audit was 9.1 hours every day, and performance in this
metric varied by sub-county, functionality, and overall maintenance condition of schemes.
Most piped schemes rely on pumping (79 percent), either wholly or partially, for distribution of
water. The main sources of energy used for pumping are generators (53 percent) and solar energy
(23 percent). Hybrid sources of energy combining a primary and a backup power source were
documented in 13 percent of the piped schemes. The team documented over 28,000 cubic meters
(m3) of installed water storage capacity in 831 storage tanks, which translates to an average storage
of two tanks per piped scheme, each with 33 m3 capacity. Overall, a metering ratio of 50 percent
was recorded for source or production points and 76 percent for connections to households,
institutions, standpipes, and kiosks. Sixty-seven percent of the piped schemes did not record meter
readings despite having a meter.
10
Ninety-five percent of the piped schemes did not treat water before distribution to users, while 21
(5 percent) indicated they practice some form of water treatment. Commonly used water-treatment
methods were chlorination (12 piped schemes) and flocculation (6 piped schemes). Users from 17
percent of all piped schemes expressed concerns over the quality of their water source. The main
issues reported relate to salinity or salty taste of water (53 percent) and assumed contamination (14
percent), from reports of stomach aches or diarrhea after drinking water from the piped scheme
source. Water supplied by piped schemes is primarily used for domestic purposes — namely
drinking, washing, bathing, and cooking — and for productive uses (brick making and irrigation).
Commercial Performance of Piped Schemes
Nine in 10 piped schemes administer a “per jerrycan” tariff, and users generally pre-pay for water at
the collection point using cash. A small proportion of piped schemes (5 percent) charge households
a monthly flat fee of $0.77 on average for unlimited water access. The average sales price for a 20–
22 liter jerrycan of water is $0.03, and credit sale of water to households is practiced in one-third of
the piped schemes. For tariff payment, 1 in 2 piped schemes reported that at least 40 percent of the
revenue is not collected from users due to non-compliance. A sizeable proportion of piped schemes
(42 percent) reported having an active bank account for purposes of revenue and savings
management, while 33 percent of the piped schemes do not bank revenues and entrust a committee
member with cash storage. Mobile banking is an emerging alternative embraced by 5 percent of the
piped schemes audited. The leading operational expenditure items are fuel or electricity (42
percent), salaries (36 percent), and repair and maintenance (13 percent).
Governance and Management
Eighty-two percent of piped schemes are managed by users or communities through elected
representatives or water management committees (WMCs). The typical size of a WMC is 10
elected members, and half of the WMCs represent up to 100 households or members. Generally,
the WMC has equal representation of men and women. Two in three piped schemes are legally
registered as self-help groups (SHGs) or community-based organizations (CBOs) with the
department for social services in Kenya, have formalized by-laws to regulate their conduct, and
regularly prepare minutes of their WMC meetings. Some of the cross-cutting management
weaknesses documented by the audit include failure to maintain technical performance records and
books of accounts, develop plans and budgets, and conduct regular and independent audits.
A total of 711 people are employed in the county to support operations of piped schemes. On
average each piped scheme employs two staff members to fill the common roles of kiosk attendant,
pump operator, or watchman. Sixty-nine percent of the scheme employees attained primary school-
level qualification, and less than 3 percent attained tertiary or college-level qualification. The average
monthly salary paid to employees was $42 and varied based on the role and level of education.
Opportunities for Rural Water Service Sustainability
Article 43 of the Constitution of Kenya (2010) entrenches water and sanitation as a right, and
Article 21 obligates county governments to take steps to ensure they progressively realize this right.
The water audit methodology provides a low-cost and adaptable approach for other county
governments in Kenya to establish a baseline status of water services and advance service provision
in their jurisdiction areas. For Kitui County, 3,126 sources spread over an area of 24,194 square
kilometers (km2) were mapped at an all-inclusive cost of $80,000.
A Water Infrastructure Audit of Kitui County 11
To better promote sustainability of rural water services, Kitui and other county governments need
to: (1) integrate climate risk analysis in the selection of water source type and location for
investment to mitigate drought risks and improve service reliability; (2) explore potential benefits of
a standardization policy for technologies deployed by actors, considering that multiple brands of
equipment have been installed over time and considering the added complexity to repair and
maintain service provision; (3) build capacity of WMCs to embrace simple but essential practices and
systems of operations management to improve performance; and (4) address the widespread land
disputes that are documented and review the traditional model by WASH actors wherein “land
contribution” by communities is linked with sustainability.
Even when all the equipment is working, the water audit shows that piped schemes close operations
in the wet season because users shift to using free surface and groundwater alternatives. The
primary alternative is shallow wells, which are an unimproved source and expose users to health
risks. Use of alternative sources in the wet season also negatively impacts business viability of piped
schemes, and there is thus a need to counter these deep-rooted behavioral issues to achieve
sustainable and universal water service coverage.
Kitui County’s water strategy aims to reduce the distance covered to water source by county
households to 2 km by 2022. The water audit provides indicative figures of the average distance
covered by households for each sub-county and accentuates inequalities. There is an opportunity for
Kitui County Government to enshrine the “distance to source” metric in relevant county policies on
resource allocation and to target resources or investments in the coming years to specifically
address inequalities.
Finally, there is an opportunity to (1) establish an information management system or database to
build on the baseline information provided by the water audit and (2) design an updating mechanism
to enable cost-effective monitoring of the county water sector. For better management and
monitoring of water infrastructure, Kitui and other county governments will also need to develop
unique identification format and physical tagging of water points.
Conclusion
The audit provides an invaluable data set of key water sources and a baseline status of water services
in Kitui County. It suggests that investments in county water infrastructure have grown considerably,
but sustainability of water services remains a challenge. Non-functionality of water sources is the
main drawback to the county’s goal of ensuring that improved water and sanitation services are
available and accessible to all by 2030, as outlined in the Kenya Vision 2030 blueprint for the water
and sanitation sector.
Commissioning a database to hold the water audit data set, coupled with an updating mechanism,
will support efficient monitoring of the county water sector. Targeted support of the rural water
sub-sector is necessary to improve performance of the popular community management model. This
may include integration of public-private partnership (PPP) models of operation and maintenance
(O&M) for rural sources to amplify performance.
County governments in Kenya are tasked by the constitution with ensuring universal water service
access, and leveraging the private sector would complement their finances, technology, and expertise
in delivery of this mandate. This will require creation of better policies and incentives to attract the
12
private sector and investment of county resources to accelerate uptake of novel PPP models at
scale.
A Water Infrastructure Audit of Kitui County 13
Introduction
Overview of Kitui County
Kitui County is one of the 47 counties of Kenya and is located about 160 kilometers (km) east of
Nairobi City. Kitui is Kenya’s sixth-largest county, covering an area of 30,496.51 square kilometers
(km2), of which 6,302.7 km2 are occupied by Tsavo East National Park. The county topography is
generally low-lying and falls within the altitude range of 400 to 1,800 meters (m) above sea level.1
Administratively, Kitui County is further sub-divided into eight sub-counties: Kitui Central, Kitui
West, Kitui East, Kitui South, Kitui Rural, Mwingi North, Mwingi Central, and Mwingi West (see
Figure 1). Each sub-county is composed of several wards, and each ward represents several villages.
There are a total of 40 wards and 246 villages in Kitui County.2 The sub-counties, wards, and villages
are governed by appointed officers of Kitui County Government, namely sub-county administrators,
ward administrators, and administrators.
Kitui County’s population was 1,012,7093 in 2009 and was composed of a 53 percent (531,427)
female and 47 percent (481,282) male population. The county’s annual population growth rate is
approximately 2.1 percent and slightly lower than the national average population growth rate of 2.6
percent. In 2018, the county’s estimated population was 1,221,001 people. Kitui Central was Kitui
County’s most densely populated sub-county, with 238 people per km2, and Kitui East was the least
populated, with 29 people per km2. The county average was 50 people per km2. The average
household size was five people per household.
1 Kitui County Government (2013), First Kitui County Integrated Development Plan 2013–2017:
Planning for Sustainable Socio-Economic Growth and Development. 2 Kitui County Villages Act 2015, Part III. Available at:
http://kenyalaw.org/kl/fileadmin/pdfdownloads/Acts/KituiCountyVillagesAct2015.pdf (Accessed June
4, 2018). 3 Kenya National Bureau of Statistics (2009), The 2009 Kenya Population and Housing Census,
Nairobi, Kenya.
Kikanga Sand Dam in Matinyani Ward, Kitui, was audited on November 21, 2017. Photo Credit: Cliff
Nyaga/University of Oxford
14
The two main urban centers in the county are Kitui Town and Mwingi Town, which together
account for 14 percent of the county population. The remaining 86 percent of the county population
is categorized as rural.
Kitui County falls in the semi-arid climatic zone of Kenya and is mostly dry and hot, with
temperatures ranging between 14°C during the coldest months (July through August) and 34°C
during the hottest months (January through March). The maximum mean annual temperature ranges
between 26°C and 34°C, whereas the minimum mean annual temperature ranges between 14°C and
22°C. The county receives between 500 and 1,050 millimeters (mm) of rainfall annually, with average
rainfall of 900 mm a year. The topography of the landscape influences the amount of rainfall received.
The highland areas of Mumoni Hills to the north, Kitui Central, Mutitu, and Endau Hills receive 500
to 1,050 mm per year, while the drier lowlands stretching from the north (Tseikuru, Kyuso, Mwingi,
Ngomeni, Nguni, and Nuu) through the Yatta plateau, the eastern areas (Mutito and Mwitika), and
southern areas (Mutomo and Ikutha) receive less than 500 mm.4 Kitui County has two rainy seasons:
May through June (long rains) and September through October (short rains). The short rains are
more reliable and are the county’s principal productive season. The long rains usually provide about
30 percent of crop production, mainly pulses, including green grams and pigeon peas.
4 Kenya Meteorological Service (2015), Kitui County Climate Information Services Strategic Plan.
Available at:
http://www.adaconsortium.org/index.php/component/k2/item/download/51_c1737bfaf1145ef5fcf15d
d148e6b742 (Accessed June 11, 2018).
Figure 1. Map of Kenya (left) and map of Kitui County showing the eight sub-counties (right)
A Water Infrastructure Audit of Kitui County 15
Devolution of Water Services in Kenya
Articles 6, 174, 175, and 176 of the Constitution of Kenya (2010) created a two-tier system of
governance comprised of one national government and 47 county governments. The responsibility
to provide water and sanitation services in Kenya is now devolved and a function of county
governments. In addition, Article 43 of the Constitution of Kenya entrenches water and sanitation as
a constitutional right by recognizing the right to “reasonable standards of sanitation” and “clean and
safe water in adequate quantities.” This obligation, as per Article 21, should be delivered by the
county governments and requires them to take steps to ensure they progressively realize this right.
Linked with this, the Government of Kenya is implementing its economic blueprint, the Kenya Vision
2030,5 which seeks to make Kenya a newly industrializing middle-income country providing high-
quality life for all its citizens by 2030. The blueprint, among other objectives, outlines the goal of
delivering universal access to water and sanitation by 2030 and identifies priority investments.
Kenya officially endorsed the sustainable
development goals (SDGs) in late 2016, at
which time it was already in the eighth year
of implementing the Kenya Vision 2030,
executed in five-year cyclic Medium-Term
Plans. By the time the SDGs were adopted,
the second Medium-Term Plan (2013–2017)
and the first County Integrated Development
Plan for the same period were in their third
year of implementation. Preparation of the
third Medium-Term Plan and the second
County Integrated Development Plan for the
period of 2018–2022 is complete, and these
plans have further mainstreamed the SDGs.
The Water Act (2016)6 provides the policy
framework for alignment of Kenya’s water
sector to the devolved structure of
government described in the Constitution of
Kenya and clarifies mandates of the various
national, regional, and local-level sector
institutions and actors. The Water Act
mandates county governments to establish
Water Service Providers (WSPs) and the
Water Services Regulatory Board (WASREB)
to license, regulate, and maintain an
information system of water service levels
5 Government of Kenya, 2007. Kenya Vision 2030: A Globally Competitive and Prosperous Kenya. 6 The Water Act, 2016, Part IV. Available at:
http://kenyalaw.org/kl/fileadmin/pdfdownloads/Acts/WaterAct_No43of2016.pdf (Accessed March 7,
2018)
Eighty-nine rock catchment sources (non-equipped)
were mapped in Kitui County. Photo Credit: Cliff
Nyaga/University of Oxford.
16
and quality, among other roles. As of 2017, 88 WSPs in Kenya were regulated by WASREB, including
two WSPs established by Kitui County Government.7
Kitui County Government owns two WSPs, Kitui Water and Sanitation Company (KITWASCO)
and Kiambere-Mwingi Water and Sanitation Company (KIMWASCO). These WSPs are licensed and
regulated by WASREB to provide water services in the county, in accordance with a set of
regulations and quality standards established by WASREB.8 However, the two WSPs currently serve
roughly one-third (390,000) of the total county population (1,221,001) due to various constraints
that include limited infrastructure coverage, inefficient operations, inadequate water resources,
urbanization, and unprecedented demand or population growth. The remainder of the county
population is partly served by rural sources or schemes that play a critical role in closing the water
access gap. These sources have been developed by different actors, including national and county
government agencies, bilateral and multilateral donors, NGOs, churches, philanthropists, and users
or communities. County-wide estimates indicate that 42 percent of Kitui County’s population has
access to at least a basic water service (i.e., time taken for a return trip to collection point is 30
minutes or less).9
Almost all small rural schemes are unlicensed and unregulated by WASREB, and availability of
information on their functionality, performance, and service quality is limited. For Kitui County
Government to develop and actualize its water service delivery goals, develop realistic investment
plans and policies, and accurately inform national-level monitoring, there is a need to shed light on
performance, impact, and the potential role of rural schemes in transforming county water service
delivery.
At the national level, WASREB is leading various initiatives to bring more rural water supply schemes
under its regulation. Key among them is a review to develop simple and relevant indicators that will
allow county governments to efficiently monitor and report to WASREB on performance of rural
schemes in their jurisdiction area. Further, in recognition of challenges with the traditional rural
water service delivery model in which operation and management of infrastructure is often left to
local communities with limited capacities, WASREB is developing public-private partnership (PPP)
models for potential adoption by county governments and sector actors to address some of the
fundamental issues and increase sustainable water service access. This may also help to bridge the
$12 billion funding gap required for Kenya to deliver universal water access by 2030.10
7 Water Services Regulatory Board (2018), Impact Report, Issue 10: A Performance Report of Kenya
Water Services Sector 2015/2016 and 2016/2017. Available at: https://wasreb.go.ke/impact-reports/
(Accessed August 1, 2018). 8 Water Services Regulatory Board (2018). Water Services Regulatory Board’s Website. Available at:
https://wasreb.go.ke/water-service-regulations/ (Accessed August 1, 2018). 9 Kenya National Bureau of Statistics (2018), 2015/2016 Kenya Integrated Household Budget Survey,
Kenya. 10 Ministry of Environment, Water and Natural Resources & JICA (2014), The Project on the
Development of the National Water Master Plan 2030.
A Water Infrastructure Audit of Kitui County 17
Kitui County Water Audit
What Is a Water Audit?
A water audit entails undertaking a comprehensive assessment of equipped sources (see Box 1)
across construction details and operation performance, including: location details, technical
specifications of parts or features, maintenance history, reliability, water quality, breakdown history,
maintenance practices, commercial performance, management, and governance structures in place.
Selected water quality parameters — namely pH, electrical conductivity, total dissolved solids (TDS),
and temperature — are measured using a probe meter. Data collected include image files of the
water source documentation, meter reading records, and physical infrastructure.
The University of Oxford developed the water audit methodology and piloted it in Mwingi North
Sub-County in August 2016 with funding from the REACH Program.11 The audit for the remaining
seven (of eight) sub-counties of Kitui was supported by the Sustainable WASH Systems Learning
Partnership (SWS), a global United States Agency for International Development (USAID)-funded
initiative to identify locally-driven solutions to the challenge of developing robust systems capable of
sustaining water, sanitation, and hygiene (WASH) services. Implementation of SWS is ongoing in
11 REACH is a global research program to improve water security for 5 million poor people in South
Asia and sub-Saharan Africa by delivering world-class science that transforms policy and practice.
https://reachwater.org.uk
Box 1: Definition of source types
Equipped sources
• Piped scheme: a water source (whether borehole, rock catchment, shallow well, earth dam,
or sand dam) incorporating a water distribution network and/or including pumps, tanks,
valves, meters, pipeline, kiosks, and standpipes, among other features.
• Hand pump: a well or borehole with a hand-operated pumping mechanism installed for
groundwater abstraction.
• Water kiosk: a WSP-operated or -managed kiosk that forms an extension of its network.
Non-equipped sources
• Borehole: a machine-drilled source that is not equipped, whether under construction,
abandoned, or stalled.
• Rock catchment: a reservoir created by storing runoff from a rock surface or catchment.
• Shallow well: a large-diameter, hand-dug hole in the ground for abstraction of groundwater.
It may include a bucket and rope mechanism for water abstraction.
• Earth dam: a reservoir created by excavating open ground to collect and store surface
runoff, with a pipe network installed to prevent direct drawing of water by people and/or
livestock.
• Spring: the natural flow of groundwater from an aquifer to earth’s surface, both protected
and unprotected.
• Sand or sub-surface dam: a small wall built above ground and into the riverbed of a seasonal
river to accumulate sand upstream and obstruct groundwater flow resulting in water storage
below ground level.
18
Kenya, Ethiopia, Uganda, and Cambodia by a consortium of eight partners led by the University of
Colorado Boulder (UCB).
In Kenya, SWS will be implemented over 5 years, from 2017 to 2021, by the University of Oxford,
UNICEF Kenya, and Rural Focus. This program of work aims to test, demonstrate, learn, and share
knowledge on the application of systems-based approaches to improving sustainability of rural water
services. The geography of focus is Kitui County, working in close partnership with the County
Ministry of Agriculture, Water, and Livestock Development (MAWL).
Goals of the Audit
The broad objective of the water audit was to conduct an objective assessment of all major water
infrastructure and provide evidence on the status of rural water services in Kitui County. The audit
also sought to inform dialogue and consensus building on issues that need to be addressed to
achieve sustainability of rural water services.
The specific objective was to locate and map all equipped rural water infrastructure and collect
information on installation details and operational performance to inform county dialogue, planning,
policies, and investment decisions at MAWL. In addition, the audit would provide the basis for the
development of a water service monitoring system for Kitui County Government. Non-equipped
water sources, defined as sources with no distribution network or pumping mechanism installed (see
Box 1), were also audited, although less extensively compared to equipped sources.
Kwa Muithya shallow well in Mutonguni Ward of Kitui was developed by community members in 2008.
The well is unreliable during the dry season due to receding of groundwater table. Photo Credit: Cliff
Nyaga/University of Oxford.
19
Approach and Methodology
County-Level Consultations
The audit scope, tools, and work plan were developed and reviewed with MAWL and adapted prior to
commencing field work to align with Kitui County Government’s plans and needs. MAWL undertook
formal introduction of the water audit activity to relevant ministries, including the County Ministry for
Administration and Coordination, and to local county government officers to encourage their
cooperation and support of the field data collection activity.
Recruitment and Training of Enumerators
A one-tier recruitment process for enumerators involved circulation of the terms of reference, with
more than 300 applicants expressing interest. The desired threshold was a water-related diploma (or
higher) qualification; prior experience in WASH-related surveys or research, preferably in Kitui; and
proficiency in the local language, Kamba. The team shortlisted 23 potential candidates and conducted
face-to-face and phone interviews. Fourteen enumerators and two12 team leaders were then selected
and trained for 5 days, November 6–10, 2017. The first three days of the training covered the audit
purpose and approach, rural water sources and their characterization, audit tool questions, their
meaning and translation to Kamba for uniformity of understanding, and the use of the adopted mobile
field data collection platform, which integrates a web (ONA) and mobile app (Open Data Kit) product.
Participants were also trained in data management and submission using the Open Data Kit app.
A pilot exercise was conducted on day four of the training to simulate the actual field data collection.
For logistical convenience, training and piloting were both done in Mwingi North (Kyuso Ward), and the
group of 16 enumerators was split into two, with eight enumerators in each group. Each group visited
three different water source types and conducted interviews with the WMCs (for piped scheme
sources) and users if available on site (for kiosks, hand pumps, and non-equipped sources). On day five,
all data collected in the pilot exercise were reviewed and several changes made to the audit instrument
to accommodate all emerging issues and enumerator feedback.
Field Data Collection
Field data collection kicked off on November 13, 2017 with three field teams of 4 to 6 enumerators
each. Team A, with six enumerators, covered Mwingi Central, Kitui East, and Kitui South Sub-Counties.
Team B, with four enumerators, covered Mwingi West and Kitui Rural Sub-Counties. Team C, with four
enumerators, covered Kitui West and Kitui Central Sub-Counties.
12 A third team leader, also part of the audit team, was a staff of the SWS program and did not require
recruitment and training.
20
Each team was headed by a team leader with a
minimum of an undergraduate degree
qualification whose daily role included logistics
management, pre-arranging audit meetings with
WMCs for piped schemes, data quality control,
including checks for completeness, and
managing online submission of completed audit
forms. A project manager was responsible for
the overall audit exercise; their primary role
was to provide technical support to field teams
if/when required and data quality control. Team
leaders pre-arranged meetings with WMCs in
close liaison with local county officers (sub-
county water officers, ward administrators, and
village administrators), who had received a
prior directive from Kitui County Government
to support the field activity. Audit meetings
would often be held at the water source site
and followed by a “transect walk” with the
WMC across the distribution network to map
and assess the condition of key infrastructure
features at the scheme in question, including
storage tanks, generators, pumps, standpipes,
valves, and kiosks.
Each enumerator was provided with a tablet
computer, a handheld GPS (where the tablet’s
GPS functionality was weak), a power bank (where the tablet battery did not last for a full day), a map of
the assigned sub-county, and stationery. Locally sourced motorcycle taxis (boda bodas) were used to
provide transport for efficient movement from one water source to another.
A daily target was set in which each enumerator was expected to audit one piped scheme and at least
five other non-equipped sources. The audit instrument for piped schemes was more detailed and
therefore longer. It required a meeting with a WMC to complete, unlike other source types. Field data
collection ended on December 23, 2017, with teams spending between 27 and 36 working days
(including Saturdays) in the field.13
13 The team that covered Mwingi West and Kitui Central Sub-Counties took 27 days to complete field
work.
WMC members for a hand pump source convene for
the water audit exercise. Photo Credit: Cliff
Nyaga/University of Oxford.
A Water Infrastructure Audit of Kitui County 21
Data Quality Control and Cleaning
Data quality checks were performed daily and at two levels. First, at the end of the day, the team leader
checked completed surveys in the Open Data Kit for any inadequacies before online submission to the
ONA server for web access. The team leader would also tally visited water sources against the
inventory provided by MAWL to ensure all known sources in a target village were located and audited.
The project manager would then access data submitted to ONA web on the previous day for further
quality checks. Figure 2 shows a sample report available on the ONA dashboard, which allowed for a
quick appraisal of teams and submitted data.
Submitted data were triangulated using Excel analysis, which the audit instrument was designed to allow.
Image files on ONA web (see Figure 3 below) were used to visually validate some of the discrepancies
noted, such as the wrong distinction between some of the rare hand pump types. A daily report was
provided to team leaders seeking clarification and/or specifying sources to be re-audited by enumerators
to fill material gaps if noted.
Figure 2. Sample report on the ONA dashboard providing a quick analysis of audit data
22
The final phase of data cleaning began immediately after completion of field data collection. In general,
this involved deleting duplicates for sources where enumerators had been sent for a repeat audit
exercise. Distinct outliers in the data set were also isolated for further verification and correction. This
process resulted in minor adjustments, especially in cases where GPS coordinates had been manually
entered due to commission errors. Eight duplicate water points were deleted from the data set in
addition to a few other corrections of individual data elements. The cleaned data were then merged
with data sets from previous audits (for Mwingi North Sub-County) to enable a comprehensive county-
wide analysis.
Water Audit Limitations
1. The primary goal of the audit was to locate and assess major water infrastructure comprising hand
pumps and piped schemes. Because non-equipped sources that skirted the hand pumps and piped
schemes were targeted and mapped, it is unlikely that all non-equipped sources were found. The
“kiosk” source type represents WSP-managed infrastructure; they were not a priority for the audit
but were included following a request from Kitui County Government to provide them with an
additional layer of information.
2. The water audit covered seven sub-counties and excluded Mwingi North Sub-County, where a
similar exercise was conducted in 2016 (for piped scheme sources) and 2011 (for hand pump
sources). Data sets collected from these different periods were updated with the available
information and merged for analysis to provide the county-wide picture included in this report. For
some of the metrics, required information was not available for updating.
Figure 3. Photo menu of ONA web showing image data submitted by enumerators
A Water Infrastructure Audit of Kitui County 23
3. The Mwingi North Sub-County data set that was collected in 2011 and 2016 and is included in this
analysis is composed of equipped sources only (i.e., piped schemes and hand pump source types).
However, for the remaining seven sub-counties, key findings on both equipped and non-equipped
sources (see Box 1) are reported.
4. The water audit data were collected in greater detail for piped schemes than any other source type,
mainly because they comprise the key infrastructure in rural geographies providing safe14 drinking
water to rural households. Piped
schemes also serve relatively more
people than point sources, and the
water audit sought to provide
empirical evidence to support
planning and investments by the
county WASH actors in
development, operation,
maintenance, and monitoring of
this type of infrastructure.
5. Information on piped schemes was
sourced primarily through
interviews with WMCs and
scheme operators and, where
possible, by inspection of scheme
records. In many cases,
construction and operations
records were either unavailable or
limited, and enumerators
frequently relied on self-reporting
by WMCs to fill gaps. Accordingly,
it is important to recognize that
this may have increased the risk of
bias or inaccuracy of some of the
responses or data collected and
reported.
14 The nature of piped scheme construction prevents outside contamination. A few of them incorporate
basic water treatment, mainly sedimentation and chlorination.
Water from piped schemes is supplied to households through
multiple collection points, mainly water kiosks and standpipes.
Photo Credit: Cliff Nyaga/University of Oxford.
24
Audit Results
Water Sources, Distribution, and Coverage
Figure 4. Distribution of the 3,126 water sources located mapped through the water audits
in 2011, 2016, and 2017. The map is not exhaustive because it is likely that more non-
equipped sources have been installed but were not located.
A Water Infrastructure Audit of Kitui County 25
The combined water audits have located and mapped 3,126 water sources to date, spread across Kitui
County (see Figure 4). Non-equipped sources are predominant, followed by equipped sources. Among
the equipped sources, hand pumps comprise the most installed infrastructure in the county at 22
percent (687), followed by piped schemes at 15 percent and water kiosks at 9 percent. The water audit
mapped 274 KIMWASCO- and KITWASCO-managed water kiosks, of which 3 in 5 are located in
Kitui’s two urban sub-counties (Mwingi Central and Kitui Central).
Across Kitui County, 454
completed piped schemes were
audited and most (357) rely on
boreholes as their water source.
Fifty distribute water from rock
catchment sources, 20 depend
on spring sources, 16 depend on
shallow wells, and 6 piped
schemes supply surface water
from earth dam sources. The
remaining 11 piped schemes get
water from multiple sources, a
combination of a sub-surface
dam (or sand dam) and
river/stream source or an earth
dam and a river/stream source.
A total of 687 hand pumps were
mapped. Of these, 82 percent
were installed on hand-dug wells
(wide diameter) while 18
percent were installed on
machine-bored boreholes.
Boreholes, 52, 2%
Earth Dams, 613, 19%
Handpumps, 687, 22%
Piped Schemes, 460, 15%
Rock Catchments, 89,
3%
Sand Dams, 655, 21%
Springs, 28, 1%
Shallow wells, 268, 8%
Kiosks, 274, 9%
Figure 5. Distribution of water sources mapped by type, number,
and proportion of the total 3,126 sources
26
Table 1. Number of water sources mapped by type, and their distribution by sub-county
Among all non-equipped sources, sand dams (21 percent) comprise the most installed infrastructure,
followed by earth dams (19 percent), shallow wells (8 percent), rock catchments (3 percent), boreholes
(2 percent), and springs (1 percent), as indicated in Figure 5.
Fifty-two non-equipped boreholes were mapped in the audit, 22 of which were under construction,
while the remaining 30 were drilled and abandoned. This occurred for various reasons but was mainly
due to a lack of funds to complete installation, low yield (inadequate source), and water quality-related
issues such as reported colored water and high salinity. Almost all springs (26 out of 28) are located in
four sub-counties of Kitui: Kitui Central, Mwingi Central, Kitui East, and Mwingi West. An additional 20
spring sources with water distribution infrastructure installed were documented but are categorized
under piped schemes in line with the definitions used in this report.
By geographical spread, Figure 6 shows that most water infrastructure is installed in Mwingi Central (17
percent), Kitui South (15 percent), and Kitui Rural (15 percent) Sub-Counties.15 The types and number
of sources in each sub-county vary widely, as detailed in
15 Data for Mwingi North was collected in 2011 and 2016 and captures information related to hand
pump and piped scheme sources only.
Sub-County/
Source Type
Kitui
Central
Kitui
East
Kitui
Rural
Kitui
South
Kitui
West
Mwingi
Central
Mwingi
North
Mwingi
West
Grand
Total
Boreholes 3 5 8 5 2 16 4 9 52
Earth dams 31 77 61 142 71 147 2 82 613
Hand pumps 71 94 114 87 81 95 78 67 687
Piped schemes 35 62 36 97 37 68 49 76 460
Rock
catchments
13 3 36
22
15 89
Sand dams 36 77 176 73 74 64
155 655
Springs 12 5 1 1
4
5 28
Shallow wells 52 41 49 22 39 51 1 13 268
Water kiosks 96 7 19 15 56 68
13 274
Total 336 381 467 478 360 535 134 435 3,126
A Water Infrastructure Audit of Kitui County 27
.
Reliability of water sources is a key issue stemming from a climate-related lack of year-round availability
of water at sources mapped. An analysis of the 1,132 dams and ponds mapped shows that 1 in 2
dams/pans do not supply water throughout the year (see Figure 7). This is more pronounced for surface
dams (rock catchments and earth dams), compared to sub-surface dams or sand dams.
11%
12%
15%15%
12%
17%
4%
14%
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
0
100
200
300
400
500
600
Kitui
Central
Kitui East Kitui
Rural
Kitui
South
Kitui
West
Mwingi
Central
Mwingi
North
Mwingi
West
Pro
rport
ion (
%)
Num
ber
of So
urc
es
Sub-Counties of Kitui
Distribution and the proportion of all water sources mapped by sub-counties
of Kitui (n=3,126)
Earth Dams
Sand Dams
Handpumps
Piped Schemes
Rock Catchments
Shallow wells
Boreholes
Springs
Kiosks
Proportion, out of
the 3,126 sources
Figure 6. Distribution of the 3,126 water sources mapped by sub-county
28
A detailed analysis of rainfall anomalies for Kitui County was performed using CHIRPS — a quasi-global
(spanning 50°S–50°N and all longitudes), high-resolution (0.05°) precipitation data set from 1981 to
present — to create rainfall trend analysis and seasonal drought patterns. Figure 7 (left) affirms that
there is a high inter-seasonal rainfall variability and extreme drought risk in the eastern and southern
parts of the county (deep brown patches). When layered with infrastructure mapped during the audit
85
330
52
562
69
313
180
338
17
150171
147
0
8760
0
100
200
300
400
500
600
All Dams Rock Catchments Earth dams Sand Dam
Is water available at the source (dam/pond) all year
round? (n=1132)
Don’t Know No No response Yes Don’t
Know
7%
No
50%
No
respo
nse
30%
Yes
13%
Is your dam/pond reliable
all year round? (n=1132)
Maps of Kitui showing rainfall variability (left) and infrastructure distribution (right), including the
longitude (x-axis) and latitude (y-axis) to indicate location.
Figure 7. Kitui seasonal rainfall anomalies (left) and distribution of infrastructure (earth dams, hand pumps, and
rock catchments) developed over time (right). Adapted from Ellen Dyer (University of Oxford),
Reach Program, 2018.
Figure 8. Reported annual reliability of surface and sub-surface dams
A Water Infrastructure Audit of Kitui County 29
(right), the analysis suggests that surface water sources (i.e., rock catchments, earth dams, and hand
pumps) have been constructed in areas of sequential dry periods. These sources are vulnerable to
ensuing climate extremes and are rendered unreliable during dry periods; it is therefore paramount for
Kitui County Government and its partners to ensure that future infrastructure investments consider
such definite drought risks.
Functionality of All Sources Mapped
The water audit sought to establish functionality of each source mapped, this being one of the key
performance indicators used in the sector to ascertain quality of the water service accessible to users.
Functionality was defined using the following criteria: (1) full functionality, (2) partial functionality for
sources that are broken but working sub-optimally, and (3) non-functional status for sources broken and
not working at the time of the audit. Figure 9 shows operation status registered on the day of the water
audit, with 60 percent of all water sources in
Kitui County fully functional and 40 percent
either partly working (15 percent) or not
working at all (25 percent).
Functionality varies by the type of source (see
Figure 10) and by administrative unit (see Figure
11). Generally, the highest functionality was
recorded for springs (89 percent), perhaps due
to the rudimentary maintenance skill required
for this source type, while a high non-
functionality was recorded for water kiosks16
(68 percent), hand pumps (55 percent), and
piped schemes (44 percent), all of which are
sources that are mechanized or incorporate
water distribution infrastructure.
16 No audit meetings were pre-arranged for WSP-managed kiosks. It is possible that the reported non-
functionality was partly associated with closure or lack of demand due to rains in the audit period.
Operational
60%Partly
operational
15%
Not
operational
25%
Figure 9. Functionality status of all water sources
mapped on the day of audit
30
Mwingi North and Kitui South Sub-Counties lead in terms of the proportion of sources that are
operational, whereas Mwingi West and Kitui Central lag behind all other sub-counties. One possible
contributing factor for Mwingi North’s accomplishment may be the FundiFix model,17 a PPP initiative led
by the University of Oxford and UNICEF Kenya in partnership with Kitui County Government to
address the functionality of rural water infrastructure. FundiFix is an insurance-based model that has
provided repair and preventive maintenance service to local communities and institutions managing
piped schemes and hand pumps since 2016. The FundiFix model works by incubating county-based
enterprises to provide the maintenance service — and in turn deliver year-round functionality — for
reliable water access to rural communities, schools, and health facilities. At scale, the model offers an
alternative pathway for Kitui and other county governments in Kenya to rapidly achieve universal
drinking water service coverage.
17 A rural PPP model for maintenance of rural water services, by the University of Oxford and UNICEF
Kenya, under incubation in Kitui County and that worked in Mwingi North at the time of the audit. For
more information: http://www.oxwater.uk/model.html or www.fundifix.co.ke.
76%
45%56%
63%76%
89%
59%
32%
15%
17%
15%
29%14%
11%
11%
16%
8%
38% 28%
6%8%
0%
28%
51%
0%
20%
40%
60%
80%
100%
120%
Earth Dams Handpumps Piped
Schemes
Rock
Catchments
Sand Dams Springs Shallow wells Water Kiosks
Functionality by water source type (n=3,126)
Operational Partly operational Not operational
Figure 10. Functionality of all water sources mapped on the day of audit by source type
A Water Infrastructure Audit of Kitui County 31
County WASH Partners and Investments
The 3,126 water sources mapped represent investments in the county water sector dating back to 1952
by a diverse range of actors comprised of government, NGOs, multilateral and bilateral donors,
churches, private philanthropists, and local communities. For 20 percent of the sources mapped,
development of infrastructure was financed through a collaborative effort of two or more organizations.
50% 64% 60% 67% 62% 60% 72% 55%
28% 26% 28% 21% 20% 28% 21% 18%
336
381
467478
360
535
134
435
0
100
200
300
400
500
600
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%K
itui C
entr
al
Kit
ui Eas
t
Kit
ui ru
ral
Kit
ui South
Kit
ui W
est
Mw
ingi
Centr
al
Mw
ingi
No
rth
Mw
ingi
West
Tota
l So
urc
es
(no)
Funct
ional
ity
(%)
Name of sub-County
Functionality of all sources by sub-county (n = 3,126)
Operational Partly operational Not operational No of sources
Figure 11. Functionality status of all water sources mapped and counted by sub-county
32
By number of sources, NGOs are the leading investor (37 percent) in Kitui County water infrastructure
development (see Figure 12). Key players include CARITAS, AMREF, World Vision, German Agro
Action, Adventist Development and Relief Agency, Anglican Development Services, Sahelian Solutions
Foundation, Africa Sand Dam Foundation,
Dorcas Aid, ACK Church, LDS Church,
Samaritans Pulse, and Action Aid. For this
analysis, churches and foundations that have
supported infrastructure development are
categorized under NGOs. Bilateral and
multilateral programs (e.g., the National
Agriculture and Livestock Extension Program)
were quoted in 11 percent of the sources
mapped and include JICA, Germany, GIZ, GTZ,
Kfw, USAID, DANIDA, the European Union,
Italy, Egypt, UK Aid, the World Bank, UNICEF,
World Food Program, and the Food and
Agriculture Organization. It is worth noting that
the analysis was based on reported details by
WMCs and site branding because records were
not available for verification in most cases. It is
therefore likely that more bilateral donors than
reported here have invested in the Kitui County water sector, as they constitute the primary source of
grants for NGOs. Communities or users and individual philanthropists have financed development of 10
percent of the sources — mainly shallow wells, earth dams, and hand pumps — in most instances with
co-funding from other partners.
National and county government agencies — including MAWL, Tanathi Water Service Board,18 Water
Sector Trust Fund, National Drought Management Authority, National Water and Pipeline
Corporation,19 a former national ministry of Arid Lands and Northern Development, and the
Constituency Development Fund — funded the construction of 23 percent of all infrastructure audited.
Investors or donors who supported infrastructure development were not known for 19 percent of the
sources, and even where this information was provided, documentary evidence was in most cases not
availed by WMCs for validation. The private sector is predominantly involved in construction of new
water infrastructure in the county through contracting and employed by the above-mentioned funders.
Private sector involvement in one-off rehabilitation and repair works was also documented, although at
a relatively small scale. An emerging thematic area for the rural sector is how to attract private sector
enterprises in the repair and maintenance space to principally increase functionality of water
infrastructure. For Kitui County, FundiFix (through a local social enterprise, Miambani) is one such
model that is already serving this market.
18 In transition to become a Water Works Development Agency in line with the Water Act (2016). 19 In transition to become a National Water Storage and Harvesting Authority in line with the Water
Act (2016).
Miltilateral
/Bilateral
11%
Community
10%
Don’t
Know
19%
Government
23%
NGO
37%
Figure 12. Kitui County investors by number of water
sources developed
A Water Infrastructure Audit of Kitui County 33
An analysis of capital investment data for both equipped and non-equipped sources dating back to 2000
shows that by number of sources, piped schemes, earth dams, and shallow wells dominate the
infrastructure developed between 2000 and 2017. The leading financier has been the government and
NGOs, with the exception of shallow wells, where communities have been more actively involved (see
Figure 13). By value of investment, Figure 14 suggests that piped schemes account for more than half (62
percent) of the total investment (unadjusted) in water infrastructure made over the same period,
followed by earth dams.20
Investment data for piped schemes were collected in two phases. In the first phase, enumerators
interviewed and examined scheme documentation with WMCs as part of field data collection for the
water audit. Gaps were identified following analysis of the data, and the second phase involved
contacting the different actors reported to have financed development of piped schemes of interest (i.e.,
NGOs, donors, government agencies, etc.) to obtain information on the accurate cost infrastructure
development. The first round yielded factual and estimated cost data for 12 and 77 piped schemes,
respectively. Follow-up with the various organizations yielded correct data for 74 piped schemes. Data
sets from both phases were consolidated to provide an accurate record of the 86 (19 percent) piped
schemes analyzed in this report.
20 Investment or construction cost data were available for 212 earth dams, and supporting documents
were unavailable for validation purposes.
12
%
23
%
5% 8
%
5%
11
%
4%
50
%
7%
5%
27
%
2%
23
%
15
%
25
%31
% 37
%
4% 8
%
14
%19
%
34
%
17
%
62
%
51
%
EA
RT
H
DA
MS
PIP
ED
SC
HE
M
ES
SH
AL
L
OW
WE
LL
S
HA
ND
P
UM
PS
SA
ND
DA
MS
Lead ing investors in earth dams, p iped schemes , sha l low wel l s ,
hand pumps, and sand dams
Bi/Multilateral Community No Data Government NGO
Figure 13. Leading investors in leading water infrastructure installed since 2000
34
The total unadjusted capital investment of the above-mentioned 86 piped schemes, developed in Kitui
County over the period of 1950–2017, is equivalent to $5.2 million.21 Figure 15 reveals that $5.08
million (unadjusted) was invested in the 82 piped schemes developed between 2000 and 2017, and 90
percent ($4.6 million) of this investment was realized from 2011 onward. This coincides with the
decentralization of water service provision mandate in Kenya.
21 1 United States dollar (USD) = 100 Kenya shillings (January 2019).
$0
$1,000,000
$2,000,000
$3,000,000
$4,000,000
$5,000,000
$6,000,000
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Cum
mula
tive
Inve
stm
ent
(USD
)
Year of Investment
Cummulative USD investment in selected water source types since 2000
Earth Dams (n=43) Handpumps (n=21)
Piped Schemes (n=82) Sand Dams (n=24)
Wells (n=44)
Figure 14. Cumulative unadjusted investment (USD) in leading source types since 2000
A Water Infrastructure Audit of Kitui County 35
Equipped Water Sources
Hand Pump Sources
Detailed information was collected for 609 out of the 687 hand pumps mapped in the audit.22 The data
suggests that the Afridev hand pump type is dominant (86 percent), followed by the India Mark II23 hand
pump type (7 percent). Variants of the Tara pump (eight in total) and the Tany pump (three in total)
were found and mapped; however, these hand pumps exhibited a high non-functionality rate of 75
percent and 100 percent, respectively. It was not possible to identify 5 percent of the hand pumps on
the day of audit24 due to missing super structure.
Most hand pumps (87 percent) are managed by user communities through WMCs and are formally
registered with the government of Kenya as community-based organizations (CBOs) or self-help groups
(SHGs). Seven percent of the hand pumps are privately owned or managed, and local institutions (i.e.,
schools and health facilities) manage 2 percent of the hand pumps mapped. Privately owned or managed
hand pumps exhibited a higher functionality measure compared to those managed by users or
communities (i.e., CBOs or SHGs) and schools (see Figure 16).
22 In total, 687 hand pumps were mapped by the combined audits. 23 This also lumps several India Mark II extra-deep well type hand pumps mapped in the audit. 24 Theft or vandalism of surface parts was the main reason for missing hand pump super structure and,
in a few instances, safe storage by community members following a breakdown.
5 6 7 10 11
13 17
23
31
44
61
84 86
-
10
20
30
40
50
60
70
80
90
100
$0
$1,000,000
$2,000,000
$3,000,000
$4,000,000
$5,000,000
$6,000,000
2002
2003
2004
2005
2007
2009
2011
2012
2013
2014
2015
2016
2017
Cum
mula
tive
no. of pip
ed s
chem
es
Cum
mula
tive
Inve
stm
ent
(USD
)
Year of Investment
Unadjusted USD investment in piped schemes (n=80)
Cumm. Investment Cumm. Piped Schemes
Figure 15. Cumulative unadjusted USD investment made in piped schemes from 2000 to 2017
36
Generally, as shown in Figure 17, pump rods (32 percent), foot valves (28 percent), and rising main pipes
(31 percent) constitute the most broken parts in non-operational hand pumps. Even when the pumping
or mechanical system is working, 2 in 3 hand pumps supply water reliably (i.e., all year round) and the
remaining 32 percent experience dry periods due to decreased groundwater flows, especially in the dry
season period from July to September (see Figure 18).
40%33%
77%
36%
19%
11%
16%
9%
41%56%
7%
55%
0%
20%
40%
60%
80%
100%
120%
CBO/SHG Other Private owner School
Funct
ional
ity
Stat
us
& L
eve
l
Management/Ownership option
Functionality of hand pumps by management model (n=600)
Operational Partly operational Not operational
Figure 16. Functionality of hand pumps across the different management options documented
A Water Infrastructure Audit of Kitui County 37
21%
24%
24%
31%
24%
27% 28%
27%
32%
25%
18%
25%
19% 2
1%
0%
5%
10%
15%
20%
25%
30%
35%
Fai
lure
rat
e (
%)
Hand pump features/parts
Poportion of non-operational hand pumps by the reported broken part on
the day of audit (n=223 )
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Pro
port
ion r
elia
ble
/unre
liable
(%
)
Month of the year
Proportion of hand pumps with reliable (or unreliable) water supply by
month, in an average year (n=195)
%Reliable %Unreliable
Figure 17. Leading cause of mechanical breakdown in non-operational hand pumps registered
Figure 18. Dry periods for the 32 percent of hand pumps that reported unreliable supply
38
Piped Scheme Sources
The water audit mapped 460 piped schemes, 454 with construction completed and six under
construction at the time of the audit. The latter have been excluded from the results.
Technical Performance
Energy sources: The majority of piped scheme sources rely on pumping (79 percent), either full or
partial, for distribution of water to users. The main sources of energy for pumping documented were
generators (53 percent) and solar energy (23 percent), as shown in Figure 19. Hybrid sources of energy
combining a primary and a backup power source were encountered in 13 percent of the piped schemes.
Over 30 generator brands and over 16 solar module brands have been installed in Kitui County, based
on documented evidence from the water audit.
For pumped and combined flow piped schemes, performance of the pumping technology adopted was
examined by juxtaposing the different options with functionality level. Figure 20 shows that hybrid-
powered piped schemes (i.e., those that use a combination of a generator with a grid or solar power
source) and grid-powered piped schemes recorded higher functionality levels than pure generator and
pure solar-powered schemes. In addition, pure solar-powered piped schemes recorded a higher
functionality rate (61 percent) compared to pure generator-powered schemes (55 percent). However,
fewer pure solar-powered piped schemes were operational compared to hybrid-powered schemes,
suggesting that augmentation of the power supply of a piped scheme by introducing an alternative or
backup power supply may have a positive outcome on functionality.
The analysis above does not attempt to simulate the implication of capital expenditure and operating
expenditure of respective energy technologies on functionality due to insufficient data. It is therefore
important to consider financial sustainability in selection of an appropriate energy source. Further
37%
21%
42%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Pum
pin
g+
Gra
vity
Gra
vity
flo
w
Pum
pin
g
Proportion of pumped, gravity, and
combined flow schemes (n=454)
53%
8%4%
11%
23%
1% 1%
0%
10%
20%
30%
40%
50%
60%
Gense
t
Gense
t+G
rid
Gense
t+Sola
r
Gri
d
Sola
r
Sola
r+G
rid
Win
dm
ill
Common sources of power for
pumped piped schemes (n=355)
Figure 19. Water distribution pressure for piped schemes (left) and
the primary energy source(s) in pumped schemes (right)
A Water Infrastructure Audit of Kitui County 39
studies are also needed to conclusively quantify influence of the different energy sources or pumping
technologies on the functionality of piped schemes.
Figure 20. Operation status of piped schemes by the energy sources used for water pumping
Water storage: Storage tanks are an important feature in a piped scheme; their purpose is to allow
low-cost water distribution by gravity, provide supply buffer in periods of peak demand, and guarantee
short-term water supply during a network outage. The 450 piped schemes with complete data available
accounted for over 28,000 cubic meters (m3) of installed storage capacity in 831 storage tanks. This
translates to an average storage of two tanks, of 33 m3 capacity each, installed in each piped scheme.
Concrete and masonry material tanks accounted for 76 percent of the total installed capacity, plastic
tanks accounted for 16 percent, and steel tanks accounted for 8 percent. In terms of volume, 90 percent
of the installed storage was in good working condition on the day of the audit, and the remaining 10
percent had fallen into disuse due to breakdown of the tank and, in limited cases, inadequate water
pressures impeding flows in some sections of the distribution network. In terms of number of tanks,
plastic or polyethylene material tanks manifested a higher failure rate compared to tanks of other
material. The leading cause of failure for plastic tanks was breaching or cracking.
Metering: Regular monitoring of water distribution in a WSP is an important element of ensuring
prudent management of the service. Information on volume of water abstracted and distributed to users
may guide WSP managers in quantifying supply and demand, capturing wastage or losses, prioritizing
investments, and improving billing practices. In this regard, the water audit located and mapped 235
production meters and 3,179 user meters (for schools, clinics, kiosks, standpipes, and households), all
installed and functional for the purposes of monitoring the water service. Overall, a metering ratio of 50
percent was recorded for production points and 76 percent for consumption points. For the latter, yard
15%11%
16%20%
8%
55%
61%
69%64%
83%
30%28%
16% 16%
8%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Genset Solar Grid power Genset+Grid Genset+SolarPro
pro
tion o
f pip
ed s
chem
es
(%)
Energy source for pumping
Functionality status of piped scheme across the main sources of energy used
(n=323)
Partly Operational Operational Not operational
40
or household connections (85 percent) and clinics (83 percent) reported the highest metering levels,
while community standpipes (19 percent) and kiosks (59 percent) were the least metered. On the day of
the water audit, WMCs in just 10 percent of the piped schemes produced a record of their monthly
meter readings, albeit with gaps, while 67 percent of the schemes did not keep a record of meter
readings at all. This implies limited management of operations data and their application in decision
making. In a rural piped scheme context, the WMC is ideally responsible for overseeing scheme
operations, including reading of meters to effectively account for water abstraction, sales, and losses;
however, the water audit findings suggest that this is not being done.
Novel technologies for monitoring water usage have emerged and give greater focus to improving
accuracy of information and automation to enable remote and real-time access to information. “Water
ATMs” are an example of such technologies and allow high-frequency monitoring of both volumes and
revenues, relevant to both rural and urban water supply contexts. A water ATM is an automated water
vending machine that works like a regular bank ATM, the main difference being that it provides water
instead of money. Water ATMs dispense water to users based on prepaid tariffs and are installed at
public water points such as kiosks or standpipes to achieve varied objectives, including accountability,
increased water sales revenue, low operation costs, reduced non-revenue water, among others. A
University of Oxford research team developed a Water Data Transmitter25 that measures volume of
water dispensed by a hand pump and transmits the data, via a global system for mobile communication
network, to a bespoke online dashboard on which a hand pump breakdown can be flagged and
functionality validated, supported by background analytics. More than 1,000 water data transmitters have
been installed in rural Africa and Asia and continue to support rural water service design and delivery to
local communities and schools. Sweetsense is an Internet of Things company that has also developed
sensor technology26 to support performance or functionality monitoring of borehole-type sources,
linked to a user interface. Such technologies may augment traditional metering technologies and provide
richer empirical evidence to support policies and design of robust institutions for rural water
sustainability.
Infrastructure performance: The audited schemes provide water services through 2,961 household
connections, 649 water kiosks, 317 connected schools, 182 standpipes, and 48 connected health
centers. For the partly-operational and non-operational piped schemes, the audit sought to establish
specific infrastructure components that were broken. Figure 21 confirms that pipe networks (17
percent), generators (13 percent), boreholes27 (12 percent), pumps (10 percent), and storage tanks (10
percent) constitute the most broken infrastructure. Other noteworthy issues that disrupt scheme
operations include theft and vandalism of generators and solar pumping equipment, as well as disputes
over land ownership.
WMCs were asked to provide details of all breakdowns experienced in the last 2 years, including
duration, repair cost, who financed the repairs, type of repair, and whether or not the issue was
25 Further information available at http://www.oxwater.uk/technology.html. 26 Further information available at http://www.sweetsensors.com/our-technology/. 27 This is general and reported where communities did not have details of the specific part that was
broken.
A Water Infrastructure Audit of Kitui County 41
resolved. Table 2 shows that 1 in 2 piped schemes had experienced at least one major breakdown
within the 2-year period of the water audit. Further analysis of the data suggested that, in terms of cost,
scheme failures caused by pump, generator, pipe network, or storage tank breakdown are among the
costliest types of failure to resolve and expose users to shocks due to relatively high volatility (standard
deviation > $400) of repair costs (see Figure 22). Again, using a threshold of $500 to differentiate
between major versus minor repair needs, an assessment of the investment required to fix defects
observed by enumerators on the day of audit indicates that 58 percent (261) of the piped schemes
required over $500 to fix outstanding defects, while 28 percent (129) of the piped schemes required
less than $500 to return to “normal” working condition. Just 14 percent (64) of the piped schemes were
in good working condition and required no rehabilitation or repairs.
0
50
100
150
200
250
300
350
400
450
No o
f Pip
ed S
chem
es
Number of piped schemes against infrastructure components
(n=406 - 428)
Broken Functional
Figure 21. Reported broken features in partly-functional and
non-functional piped schemes on the day of the water audit
42
Table 2. Number of major breakdowns recorded in piped schemes in 2 years
Breakdowns in the Last 2 years Number of Piped Schemes Total Breakdowns
0 160 0
1 195 195
2 40 80
3 7 21
4 3 12
7 1 7
Piped schemes (n) 406
Mean breakdowns in 2 years 0.8
394
-
177
-
1,279
438
-
290
470
1,867
148
429
85
$0
$200
$400
$600
$800
$1,000
$1,200
$1,400
$1,600
$1,800
$2,000
$0
$5,000
$10,000
$15,000
$20,000
$25,000
$30,000
Std D
ev.
(Ksh
)
Tota
l C
ost
of R
epai
rs
Major piped scheme failures reported in 2 years
Total USD cost and standard deviation for major repairs completed in 2 years
(n=82)
Sum of Cost of Repair
STD Dev
Figure 22. Reported cost of major repairs completed in 2 years and the volatility of repair costs
A Water Infrastructure Audit of Kitui County 43
Repair duration and maintenance practices: Nine in 10 piped schemes do not store spare parts
for regular repair and maintenance; those that do typically stock pipes. Overall, Kitui County
Government (both the sub-county and head office) is the most active entity involved in fixing
breakdowns in 32 percent of the piped schemes, followed by local technicians in 30 percent of the piped
schemes28 (see Figure 23). In terms of the magnitude of breakdown, the county government is typically
active in fixing major breakdowns, while local private technicians and scheme employees are commonly
active in the repair of minor breakdowns. The rule of the thumb in the sector has been that tariffs
charged by rural schemes should at least cover the costs of system repair and maintenance while
government complements this with support for assets replacement and network extension. However,
government activity was also documented in minor repairs, although marginally, in 1 in 5 piped schemes.
Based on self-reported timelines, 1 in 4 breakdowns experienced in the 2 years of the audit were
resolved within 7 days. It took up to 30 days to resolve 46 percent of all breakdowns reported by piped
schemes (see Figure 24). The downtime varied based on the type of failure; generally, pipe network and
vandalism-related failures took a shorter time to fix (169 and 159 days, respectively, on average)
compared to pump failures, which took the longest time to fix (352 days on average).
28 Analysis is based on 250 piped schemes for which the most recent infrastructure breakdown had been
resolved.
23%
5%
4%
30%
1%
1%
19%
3%
1%
1%
9%
1%
1%
0% 5% 10% 15% 20% 25% 30% 35%
County Govt - Water Dept
NGO/Donor
National government
Local technician
Local technician, sub-County govt
MCA/Politican
Other
Scheme plumber
Scheme plumber, County govt
Scheme plumber, Local tech
sub-County govt
Constituency dev. fund
Water utility
Who s
olv
ed t
he las
t bre
akdow
n
No. of piped schemes (n=159)
Figure 23. Actors involved in fixing the most recent breakdown in piped schemes
44
The audit also sought to establish whether preventive maintenance of infrastructure is a norm for rural
piped schemes, and the results show that 4 in 5 piped schemes regularly carry out at least one
preventive maintenance task. Figure 25 suggests that commonly performed tasks are pipeline inspection
for bursts or leaks (61 percent) and storage tank inspection (63 percent). Tasks performed by fewer
schemes include monitoring pump meters or gauges, pump maintenance, solar equipment cleaning, and
generator maintenance.
25%
21%
14%12%
10%
19%
0%
5%
10%
15%
20%
25%
30%
1 to 7 8 to 30 31 to 90 91 to 180 181 to 365 >365
Pro
port
ion o
f fa
ilure
s fixed (
%)
Number of Days
Downtime (days) for all failures reported by piped schemes in the 2 years to the
audit (n=240)
All failures
21%8%
23%14% 12%
61%
38%
63%
0%
10%
20%
30%
40%
50%
60%
70%
Gense
t Se
rvic
ing
Sola
r C
lean
ing
Inpect
ing
Pum
phouse
for
defe
cts
Pum
p s
erv
icin
g
Read
ing
pum
p
gauge
s/m
ete
rs
Pip
elin
e
insp
ect
ion
Mete
rs
insp
ect
ion
Tan
ks
insp
ect
ion
Proportion of piped schemes by maintenance tasks done (n=451)
Figure 24. Downtime for all failures reported by piped schemes during the audit
Figure 25. Common preventive maintenance tasks done in rural piped schemes
A Water Infrastructure Audit of Kitui County 45
Reliability of water service: One in 5 (23 percent) piped schemes close operations at certain times
of the year, mainly due to low user demand in the rainy season. Availability of free alternative sources of
water in the rainy season causes a slump in scheme revenues; respondents linked this to schemes’
inability to finance operations in such periods (see Figure 26).
For this analysis, reliability was defined as the average hours in a day that water was available for
collection at yard taps, kiosks, or standpipes when there was no breakdown. For greater accuracy, only
the data for fully operational schemes were analyzed. Reliability was derived from the number of days
water was supplied in a week and the average daily hours of water service reported by WMCs. Results
show that fully operational piped schemes supplied water for 9.1 hours every day at the time of the
audit. Performance in this metric varied by sub-county: Kitui East (10.2 hours per day) and Kitui Central
(10.7 hours per day) recorded the highest reliability, while Kitui South (8.0 hours per day) and Kitui
West (8.3 hours per day) had the least-reliable water service in the county. These figures are unadjusted
for close-downs during the wet season.
41%
19%
0%
9%
9%
0% 5% 10% 15% 20% 25% 30% 35% 40% 45%
Low demand
Lack of resources
Operator leave
Cant Afford to run Scheme
Other reason
Porportion of piped schemes that reported seasonal close-downs (n=93)
Lack of revenue
Figure 26. Leading reasons for scheme close-down at certain times of the year
46
Juxtaposing reliability with the condition
of piped scheme, as determined by
enumerators, for fully operational piped
schemes only, Figure 27 suggests that
newly constructed and well-maintained
working piped schemes provided a more
reliable service than poorly maintained
working schemes, which may underscore
the importance of maintenance in
improving quality of water services. This
analysis does not incorporate correlated
service quality indicators — such as per
capita water access — due to insufficient
data, and therefore further studies are
necessary to corroborate this inference.
Distance to piped scheme sources:
The audit sought to measure progress made by Kitui County toward achieving its goal of ensuring that
all county households have access to drinking water within 2 km. The distances reported here are not
actual or observed but rather estimated and provided by WMCs during the water audit. Results show
that at the time of the audit no sub-county had achieved Kitui County Government’s goal of providing
drinking water service within 2 km of households (see Figure 28, left). Analysis of the maximum one-way
distance travelled by households to piped scheme sources showed that the county average ranges
between 6.4 km (if we include non-operational schemes) and 6.9 km (fully operational schemes only).
There is wide intra-county variance, and the reported distance to source is lowest for Kitui West (3.4
km) and Kitui Central (3.2 km) households and above the county average in Mwingi Central (9.6 km),
Kitui South (8.1 km), and Kitui East (7.4 km) Sub-Counties. Figure 28 (right) provides the density of
piped scheme investments mapped in each sub-county, per capita and per land area, in km2. The figure
implies that the foremost sub-counties (Kitui Central, Mwingi West, and Kitui West) are the smallest in
terms of land mass and yet have the highest density of piped scheme investments per km2.
Future investments in Kitui will need to advance equity to reduce the distance travelled to the source by
households in the more expansive and sparsely populated sub-counties of Kitui. Furthermore, despite
Mwingi West Sub-County recording the highest density of piped scheme investments both per capita
and per km2, the low functionality of piped schemes (48 percent fully operational) eroded progress and
meant the sub-county ranked third, with households covering 4.8 km to the source on average. This
suggests that investment in restoration of broken piped schemes may potentially reduce the distance
travelled by users to the source.
8.8 9.3
7.3
-
2.0
4.0
6.0
8.0
10.0
12.0
New schemes Well maintaned
schemes
Poorly maintained
schemes
Ave
rage
dai
ly h
ours
of su
pply
Maintenance condition operational piped schemes
Average daily hours of water supply for
operational piped schemes, by condition
(n=253)
n=29 n=24 n=200
Figure 27. Reliability against condition of piped scheme
A Water Infrastructure Audit of Kitui County 47
Water quality and uses: Of the 454 piped schemes audited, 431 (95 percent) did not treat water
before distribution to users, whereas 21 (5 percent) indicated that they practice some form of water
treatment. Commonly used water-treatment methods were chlorination (12 piped schemes) and
flocculation (6 piped schemes).29
Seventeen percent (77 piped schemes) expressed concerns over the quality of their source. The primary
water quality issues reported relate to salinity or salty taste of water (53 percent) and assumed
contamination from reports of stomach aches or diarrhea by users after drinking water from their
respective piped scheme source (14 percent). Observed teeth discoloration (1 percent) and animal
droppings in water (1 percent) were also mentioned, while 30 percent of the piped schemes did not
provide an explicit reason for their concerns.
29 It was not possible to ascertain the treatment method, if any, used by three piped schemes.
-
2.0
4.0
6.0
8.0
10.0
Kitui
Centra
l
Kitui
East
Kitui
Rural
Kitui
South
Kitui
West
Mwingi
Centra
l
Mwingi
North
Mwingi
West
Grand
Total
Maximum (one-way) distance (Km) covered to
Source
County Target
-
0.00010
0.00020
0.00030
0.00040
0.00050
0.00060
0.00070
0.00080
-
0.010
0.020
0.030
0.040
0.050
0.060
0.070
0.080
Kit
ui C
entr
al
Kit
ui Eas
t
Kit
ui R
ura
l
Kit
ui South
Kit
ui W
est
Mw
ingi
Centr
al
Mw
ingi
No
rth
Mw
ingi
West
Density of piped schemes by sub-county
All PS per KM2
Working PS per KM2
All PS per Capita
Working PS per Capita
Figure 28. Maximum one-way distance (km) travelled by households to piped scheme
sources (left) and density of piped scheme investments per capita and per km2 (right)
48
The enumerators tested the water quality in piped
schemes using portable handheld meters, where
schemes were operational and a water sample
available. The following parameters were checked:
electrical conductivity (mS), TDS (ppm), pH, and
temperature (see Box 2).
Results showed that high salinity is widespread and
more profound in borehole sources. The mean TDS
for 223 piped schemes where a sample was available
was 1,190 ppm. TDS values above the threshold
(1,500 mg/l) set by the Kenya Bureau of Standards
and WASREB guidelines30 were recorded in Mwingi
North, Mwingi Central, Mwingi West, and Kitui East
Sub-Counties (see Figure 29). The water pH in 51
(24 percent) of 211 measured piped schemes fell
outside the 6.5 to 8.5 range set by the World
Health Organization and Kenya Bureau of Standards
for drinking water.
The water audit provides clues on localities where
salinity and other quality-related issues are likely to
occur; however, further studies are necessary to
spatially map out the geophysical and
hydrogeological properties of the water resource in
high resolution to support the delivery of safe
drinking water.
30 Water Services Regulatory Board (2008) Guidelines on Drinking Water Quality and Effluent Monitoring.
Available at: https://wasreb.go.ke/downloads/Water_Quality_&_Effluent_Monitoring_Guidelines.pdf.
Box 2: Water quality parameters tested
Electrical conductivity: a measure of the
saltiness of water done on a scale from 0 to
10,000 microsiemens per centimeter (uS/cm).
TDS: a general indicator of water quality that
measures the combined total of organic and
inorganic substances that are dissolved in
drinking water, in parts per million (ppm). EC
meters that also read TDS use a correction
factor that varies between 0.5 and 0.8 to
convert a quantity expressed in uS into an
equivalent TDS value expressed in ppm.
Temperature: EC of drinking water changes
with its hotness or coldness, and a higher
temperature will generally give a higher EC
reading.
pH: a measure of the relative amount of free
hydrogen and hydroxyl ions to determine how
acidic (pH of less than 7) or basic (pH of
greater than 7) drinking water is.
A Water Infrastructure Audit of Kitui County 49
In 4 of 5 schemes, water is used for domestic purposes (drinking, cooking, washing, and bathing) and
livestock watering. Productive uses, specifically brick making (42 percent) and irrigation (30 percent),
were also recorded in numerous piped schemes (see Figure 30) and form a significant part of rural
water use in Kitui County. These productive uses were consistently recorded across all source types,
water qualities, and tariffs. In one piped scheme, water was used for fish farming, in addition to domestic
uses.
Because of Kitui County’s semi-arid climate, low water service coverage in rural areas, and water-
intensiveness of the productive uses documented by the audit, further studies are needed to quantify the
0
2000
4000
6000
8000
10000
12000
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Kit
ui
centr
al
Kit
ui eas
t
Kit
ui
rura
l
Kit
ui
south
Kit
ui
west
Mw
ingi
centr
al
Mw
ingi
nort
h
Mw
ingi
west
TD
S (p
pm
)
TD
S (p
pm
)
sub-County
TDS (ppm) levels recorded in piped schemes by sub-county (n=223)
85%
84%
91%
91%
83%
42%
30%
3%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Drinking
Cooking
Bathing
Washing
Livestock
Brick making
Irrigation
Other
Common uses of water supplied by the piped schemes (n=460)
Figure 29. Variation of TDS (ppm) levels across the eight sub-counties of Kitui
Figure 30. Domestic and productive water uses recorded in the audit
50
impact and trade-offs between domestic and productive water uses. As highlighted in the previous
section, limited metering and record keeping made it difficult to analyze and quantify the volumes and
value of water supplied for domestic uses, irrigation, brick making, and fish farming.
Commercial Performance
Tariffs and revenues management: Nine in 10 piped schemes administer a “per jerrycan” (most
jerrycans have a 20–22 liter capacity) charge for water sold, and users typically pre-pay using cash at the
drawing point. For livestock, the same tariff is applied and a livestock unit (goat or cattle) is often the
equivalent of a jerrycan. The average sales price for a jerrycan of water is $0.03 ($0.029 for members31
and $0.031 for non-members). A few piped schemes (5 percent) charge users a monthly flat fee of
$0.77, on average, for unlimited water access.
Based on complete and self-reported data of 216 piped schemes, representing over $60,000 worth of
monthly expenses, fuel and electricity (42 percent) and salary (36 percent) costs constitute the bulk of
rural piped schemes’ expenditures (see
Figure 31). Repair and maintenance is a
distant third at 13 percent. The 216
schemes reported capital expenditures
equivalent to $165,000 in the previous
year (2016–2017) spent on pipeline
extensions (55 percent), major repairs
(42 percent), and loan repayments (3
percent).
Reported revenue collection efficiency by
WMCs was relatively high, and 1 in 2
schemes reported collecting over 60
percent of its billed income (see Figure
32). However, credit sale of water to
households is common, as affirmed by
one-third of the piped schemes audited.
A sizeable proportion (42 percent) of
piped schemes reported having an active bank account for the purposes of revenue and savings
management and stated monthly banking of collected cash as their preferred frequency. The total savings
held in bank accounts of all 112 piped schemes that made regular savings amounted to $78,000. Notably,
one-third (33 percent) of the WMCs reported that they entrust a committee member with cash
storage, locally referred to as “mattress banking.” Mobile banking (Mpesa/Mshwari) is an emerging
savings alternative that had been embraced by 5 percent of WMCs at the time of the audit.
31 Members contributed in-kind or financially toward establishment of the SHG or CBO or toward the
development of the water source. They have voting rights, may be entitled to lower tariffs, and, in some
schemes, collect dividends.
Salaries
36%
Repairs
13%Admin
3%Allowances
5%
Fuel
42%
USD 60,000 in monthly operating
expenditures (n=216)
Figure 31. Reported key monthly expense items
in piped schemes by value
A Water Infrastructure Audit of Kitui County 51
Maintaining books of accounts, preparing budgets, and conducting regular and independent audits are
not common practices. Just 1 in 10 piped schemes attempt to audit their records annually, and this was
often done informally, by fellow WMC members. The observed poor accounting for income and
expenses remains a key limitation to independent and sound analysis to ascertain scheme and sector-
level financial performance and guide sustainability measures.
Governance and Management
Nearly all (82 percent) piped schemes are
managed and operated by the beneficiary
communities, through elected
representatives or WMCs, usually set up
for each scheme (see Figure 33). The
remainder of the schemes are managed by
local institutions that include schools (8
percent), local NGOs and churches, WSPs,
and health facilities. In institution-operated
schemes, the management running the
respective institution doubles as the
scheme management. Private individuals in
Kitui County manage 4 percent of the
piped schemes, which is a significant
contribution to rural water service
provision in the county.
26%
24%
24%
13%
12%
0% 5% 10% 15% 20% 25% 30%
81-100%
61-80%
41-60%
21-40%
0-20%
Coll.
effic
iency
ban
d
Proportion of piped schemes (n=337)
Community
82%
Hospital
1%
NGO/Church
2%
Individual
4%
School
8% WSP
2%
County
Govt
0%
Who manages piped schemes? (n=452)
Figure 32. Reported collection efficiency range by piped schemes for billed revenues
Figure 33. Documented management
options in piped schemes
52
Two in three community-run piped schemes are formally registered as SHGs or CBOs32 with the
department of social services in Kenya, are guided by by-laws, and regularly prepare minutes of their
committee meetings. The typical size of a WMC is 10 elected members who represent the interests of
community members or shareholders. Generally, there is equal representation of men and women in
the WMCs for piped schemes. In operational schemes, the sitting committee had been in office for a
mean duration of 4 years at the time of the audit. One in three WMCs reported they had met once or
twice in the previous year to deliberate scheme management matters.
Figure 34 provides a distribution of all employees working in the audited piped schemes and suggests
there are at least 711 people directly employed in the schemes, the majority in community-managed
piped schemes (600), followed by schools-managed piped schemes (52).
Of the 711 employees documented, the leading employee roles observed were kiosk attendants (59
percent), pump operators (12 percent), and watchmen (12 percent), as shown in Figure 35. The average
number of employees in a piped scheme was two (range 0–20), and one employee would often assume
more than one role for small schemes. Additional roles were documented in bigger schemes, including
that of plumber, scheme manager, revenue collector, meter reader, and pipeline inspector or patroller.
The highest qualification attained by most of the scheme employees (69 percent) was primary-level
education, and less than 3 percent held tertiary/college-level or higher qualifications. This suggests a
limited ability of the rural water sector to attract highly trained professionals.
32 A legally registered group or association with membership drawn from the local level and established
to solve issues within a community. CBOs and SHGs are formally registered by the Government of
Kenya through its Ministry for Gender, Children, and Social Development.
Community , 600
Hospitals/Clinics, 3
NGOs, 7
Private
owner, 31
Schools, 52WSPs, 17
Kitui County, 1
Number of piped scheme employees (n=452)
Figure 34. Number of employees by the employing entity
A Water Infrastructure Audit of Kitui County 53
The average monthly salary in the county piped schemes and for staff holding the aforementioned roles
was $42 per month, based on available data for 521 employees. This varies ($30–$100) depending on
position or role (see Figure 36). Furthermore, higher-educated scheme employees earned a higher
monthly salary on average, though still significantly lower than the set minimum wages for the relevant
occupation and skill category in Kenya,33 of between $82 per month (for a night watchman) and $133
per month (for an unskilled artisan), as shown in Figure 37.
Figure 35. Proportion of piped scheme employees occupying observed roles
33 The Labour Institutions Act: The Regulation of Wages (General) (Amendment) Order, 2017. Available at:
http://kenyalaw.org/kl/fileadmin/pdfdownloads/LegalNotices/2017/LN111_2017.pdf (Accessed June 4,
2018).
59%
1%
2%
3%
1%
3%
12%
0%
12%
2%
5%
0%
1%
0% 10% 20% 30% 40% 50% 60% 70%
Kiosk attendant
Kiosk attendant/watchman
Line patroller
Manager
Meter reader
Plumber
Pump operator
Treasurer
Watchman
Pump operator/Watchman
Kiosk attendant/Pump operator
Secretary
Revenue collector
Proportion of employees by observed roles (n = 521)
54
The typical layout of piped scheme infrastructure comprises a source, pumping equipment, pipe network
and its accessories, storage tanks, and water collection points. The water audit sought to establish
ownership of land on which these piped scheme features are built. In 55 percent of the piped schemes,
assets have been built on CBO- or SHG-owned land, while in 21 percent of the schemes, land is wholly
or partially provided by private individuals through formal or informal agreements. Regardless, these
agreements are weak, given the widespread disputes documented during the audit between the WMCs
40
48
57 60
$0
$10
$20
$30
$40
$50
$60
$70
Primary Secondary Tertiary University
Month
ly a
vera
ge s
alar
y, U
SD
Highest education qualification attained
Mean monthly salary (USD) by education level (n=521)
$30 $39
$57
$46 $
60
$31
$59
$48
$30 $40 $46
$50
$100
$50
-
20
40
60
80
100
120
Fund c
olle
cto
r/w
atch
man
Kio
sk a
ttendan
t
Kio
sk a
ttendan
t/w
atch
man
Lin
e p
atro
ller
Man
ager
Mete
r re
ader
Plu
mber
Pum
p o
pera
tor
Tre
asure
r
Wat
chm
an
Pum
p o
pera
tor/
Wat
chm
an
Kio
sk a
ttendan
t/Pum
p
opera
tor
Secr
eta
ry
Reve
nue c
olle
ctor
Mean monthly salary by employee role (n=521)
Figure 36. Mean monthly salary reported by job/role in piped schemes
Figure 37. Mean monthly salary (USD) for piped scheme staff against education attained
A Water Infrastructure Audit of Kitui County 55
and the land owners. In many of the noted disputes, post-agreement demands for further financial or in-
kind compensation have ensued and affected service delivery.
The commonplace approach adopted for many years by donor and NGO projects in Kenya was that the
beneficiary communities would contribute land and, at times, free labor during infrastructure
construction, while the donor or NGO would provide construction materials and technical expertise.
As a result, in instances where the community did not own any land, an individual would be approached
to “contribute” part of their land in return for free water for the life of the piped scheme. However, this
approach appears to have incentivized context-inappropriate practices such as the use of water for
irrigation, brick making, water resale, and non-payment of tariffs by the land owners and usually their
extended family members, among other malpractices. Some of the reported implications included loss of
scheme revenues and lack of resources for repair and maintenance. In extreme cases, violence led to
closure of schemes and denied affected communities access to the water service.
56
Opportunities for Sustainability of Rural Water Services
Upscaling of the Water Audit Methodology
The County Governments Act of 2012, section 104, obligates the 47 county governments in Kenya to
regularly develop integrated plans containing their medium-term priorities and comprising a 5-year
County Integrated Development Plan, 10-year sectoral plan, and a Geographic Information System-based
10-year spatial plan. The first integrated plans spanning 2013–2017 were developed by counties and
adopted toward the end of 2013. In the second round, counties sought to align plans for 2018–2022
with the Kenya Vision 2030 blueprint and the SDGs. The development process for the current
integrated plans was completed in 2018.
The water audit methodology provides an approach for Kitui and other county governments in Kenya to
inform the integrated plans development process. For WASH planning, a water audit is key to establish
the correct water service baselines and provide a basis to continually monitor implementation progress
of the various county strategy documents against goals. The information can also guide counties to set
realistic objectives in their future planning. The audit methodology, in Annex A, advocates for low-cost
mobile survey solutions for data collection and management and is easily adaptable to other counties.
The audit instrument (available upon request) is also adaptable to varying geographies with flexibility to
embody local dynamics relating to water services.
For Kitui, 3,126 water sources spread over an area of 24,194 km2 (excluding 6,302 km2 county land in
Tsavo National Park) were mapped over a cumulative period of 4 months (including preliminaries) with
a combined team of 22 enumerators and at an all-inclusive cost of $80,000. The benefits of information
Water meter for the borehole at Skyokithumbi Farmers piped scheme in Kwamutonga/Kithumula
Ward, Kitui. Photo Credit: Cliff Nyaga/University of Oxford.
A Water Infrastructure Audit of Kitui County 57
collected through the exercise unequivocally exceed the cost, and therefore an opportunity exists for
both national and decentralized governments to adopt, adapt, and scale up the water audit methodology.
Identification Tagging of Water Sources
The audit data set provides location information for each water source or point mapped based on the
local administrative units (including sub-county, ward level, and village level details), local name, and GPS
coordinates. A notable issue from this approach is the similarity of local names for different and even
same water source types at the lowest administrative level (same sub-location and village). For clustered
sources, e.g., those less than 10–15 m apart34 in the case of the subject audit, the GPS positions would
overlap when plotted on a map and would make on-site identification onerous, depending on the
measurement error of GPS technology used for mapping.
A unique identification format to identify each water source is proposed, which will require a county
policy on (1) a standard format for unique identification of sources, (2) physical tagging of all water
sources following the agreed identification formation, and (3) setup of an off-site catalogue to manage
water source or point information, including their assigned identity. In the subject water audit, the
source coding format “sub-county name/ward name/village name/source type initials/count, in terms of number
of sources mapped” was used and provides a starting point for the dialogue between Kitui or other
counties and their WASH actors to develop an appropriate identification policy for water sources.
Improving Operational Sustainability of Rural Water Sources
Climate risk analysis for Kitui County suggests a high intra-county variability of seasonal rainfall, with
extreme dry events more pronounced in some parts of the county than others. Coincidentally, most of
the surface water sources (rock catchments, hand pumps, and earth dams) have been built in these high-
risk areas, which exposes them to dry periods and undermines their reliability. In the future, integration
of climate risk analysis in selection of water source type and location for investment is recommended to
mitigate this aspect and deliver a more reliable water service.
Presence of multiple brands of installed equipment (e.g., in the case of generators and solar pumping
systems) suggests complexity of supply chains for parts and expertise needed to keep these diverse
technologies properly maintained and working in rural and remote contexts. A county dialogue on
potential benefits of a standardization policy for technologies installed by actors in relation to future
repair and maintenance burden of the infrastructure is paramount.
Limited record keeping and absence of or non-adherence to operation and management best practices
in piped schemes — such as stocking of spares, convening of regular WMC meetings, service monitoring
and meter reading, banking and accounting for revenues, and conducting annual and independent audits
— was noted and ultimately weakens accountability and governance systems required for good
performance of schemes and, eventually, the quality of water service provided. There is an opportunity
to continuously build the capacity of WMCs to embrace simple but essential practices and systems of
operation and management, which may increase sustainability of the rural water service in Kitui.
34 The average accuracy range recorded by tablet computers used in the audit.
58
The water audit recorded many unresolved issues in piped schemes where infrastructure was
constructed on private land. The widespread land disputes call for a rethink of the traditional model by
WASH actors, where “land contribution” by communities for water infrastructure development has
been advocated and is assumed to contribute to sustainability of the service.
Improving Consistent Use of “Safe” and Reliable Sources
Leading causes of piped scheme closure identified in this audit report are (1) mechanical breakdown of
equipment or network and (2) the wet season, which leads to low user demand for paid water due to
the availability of surface and groundwater alternatives. As shown in Figure 38, unimproved sources
(shallow wells) constitute the immediate alternative source of water for the majority of piped scheme
users, despite their high vulnerability to contamination, as well as droughts and associated health risks.
This suggests that, even when piped schemes are working, users may still switch to unimproved
alternative water sources, especially during the wet season, so that they do not pay tariffs. Other than
the adverse effects on household health and productivity, the shift to use of such alternative sources
also affects scale of operations and impacts O&M viability because many piped schemes are not able to
collect adequate revenues to cover their operation costs and end up closing operations during the wet
season. Therefore, to sustainably advance the provision of universal, safe, and reliable water services in
rural Kitui County and other similar contexts, system approaches that not only seek to deliver 24/7
functionality of infrastructure but also seek to understand and address local practices and behavioral
drawbacks are required.
15%
13%
21%
26%
13%
7%
36%
41%
P R IMARY SECONDARY
Earth Dam Piped Scheme Spring Shallow well
Proportion of piped schemes by alternative source used
Figure 38. Reported primary and secondary alternative
sources of drinking water for piped scheme users
A Water Infrastructure Audit of Kitui County 59
Advancing Policy on Ensuring Water Access within 2 km of Every County
Household
The draft Kitui County Water Services Strategy aims to reduce the distance covered by households to a
water source to 2 km on average by 2022. The water audit results show that good progress has been
made, yet inequalities between sub-counties remain. The audit has provided an indicative maximum
distance covered by households to a water source in all eight sub-counties as of 2017 (2016 for Mwingi
North Sub-County), which allows the Kitui County Government to target its investments in the coming
years to specifically address this emerging issue. This will require a delicate balance to achieve both
equality and equity, as the results show that the smallest sub-counties have a higher density of piped
schemes per km2 compared to the more expansive, rural, and sparsely populated sub-counties of Kitui.
This has partly contributed to the latter lagging behind the former in terms of distances travelled by
household to source. Therefore, enshrinement of the “distance to source” metric in the relevant county
policies as part of the criteria for allocation of county resources will be a critical first step to address
this inequality.
The audit results also show that ensuring all piped schemes are working would potentially reduce the
average distance to source. As such, policies that ensure that all water infrastructures continue to
operate many years after construction may potentially insulate Kitui and other counties from non-
functionality-related erosion of water service coverage.
Enabling Rural PPPs for Water Services Delivery
The predominant community (CBO or SHG) model of operating and managing rural water
infrastructure in the county has not delivered full functionality, especially for the more complex sources
(hand pumps and piped schemes). Exploring integration of the community management model with
targeted PPPs may strengthen the capacity to resolve the high non-functionality and other issues
highlighted by the audit and may enable better sustainability of the sector. WMCs in 79 percent of the
piped schemes confirmed interest in FundiFix’s maintenance service, which implies demand for PPPs.
The FundiFix model is an example of a rural PPP based on the insurance logic that is advancing four
pillars for drinking water sustainability: (1) monitoring, (2) professional maintenance service provision at
scale, by incubating a local social enterprise, (3) sustainable finance, and (4) coordination by government.
In the FundiFix model, the community (through WMCs) and local institutions managing rural water
infrastructure enter into performance-based contracts with a locally established FundiFix enterprise for
guaranteed maintenance and repair of broken water infrastructure within 3–5 days. FundiFix has been
working in Kitui County though a local enterprise, Miambani, since mid-2016 and has seen increased
functionality of piped schemes and hand pumps. Disaggregating functionality by sub-counties is one
indication; for example, better performance of piped schemes (69 percent) was recorded for Mwingi
North (see Figure 39), coinciding with FundiFix’s presence in the sub-county. The FundiFix model has
guaranteed reliable water services for 57,000 people in Mwingi North thus far, including schools and
health facilities.
At the current scale of FundiFix’s operations in Kitui County, community or user payments are
insufficient to cover local costs of maintenance service provision. For the last 2 years, FundiFix has been
60
underwritten by a Kitui County Water Services Maintenance Trust Fund that is part of a
learning/research activity. Coincidentally, Kitui County’s Water Strategy seeks to improve service
delivery and sustainability of installed water infrastructure by reducing the repair response time from 3
months to 1 week on average by 2022. FundiFix’s results demonstrate an opportunity for the county to
build partnerships with the private sector and expedite the realization of this objective.
In addition to the FundiFix model, different private-sector-based water service delivery models aimed at
improving performance and quality of water service in rural and low-income areas have been developed
and piloted in Kenya and other countries. These include private operator/management contracts, lease
contracts, professionalized management, and delegated management contracts, among others.35 Kitui
and other county governments could consider these models to strengthen rural water service delivery.
This would allow them to strengthen focus on policy development, monitoring and regulation, planning,
and coordination roles. Engaging the private sector would also contribute to the financing strategy
adopted by many county governments of mobilizing the private sector to realize their development
targets. Accordingly, to advance universal access to safe and affordable drinking water service in rural
Kenya, a conversation is needed to (1) ascertain which PPP models are appropriate for interested
counties, (2) develop required policies and institutions to support such models, and (3) allocate smart
investments or incentives for adoption of the selected models at scale and in return for impact.
35 Water Sector Trust Fund (2017), Service Delivery Model Toolkit for Sustainable Water Supply.
Available at: https://waterfund.go.ke/publications (Accessed August 23, 2018).
49%57% 53% 57% 59% 65% 69%
48%
14%15% 19% 10%
14%11% 8%
26%
37%28% 28% 33% 27% 24% 22% 26%
0%
20%
40%
60%
80%
100%
120%
Kitui
Centr
al
Kit
ui Eas
t
Kitui
Rura
l
Kitui
South
Kitui
West
Mw
ingi
Centr
al
Mw
ingi
Nort
h
Mw
ingi
West
Piped schemes functionality by sub-county (n=454)
Operational Partly-Operational Not Operational
Figure 39. Functionality of piped schemes by sub-county
A Water Infrastructure Audit of Kitui County 61
Water Services Information Management System
This report provides a baseline status of county water sector performance as of December 2017;
however, there is currently no systematic mechanism in place to capture and update most changes that
have happened since. For that reason, audits provide a solid basis for county governments to
commission water services monitoring systems or databases to resolve this shortcoming. A database
that is linked to an updating mechanism may ensure that the audit data set remains up to date in the
future and provide a more efficient tool for progressive tracking of changes in the sector. A simple
inaugural database capable of managing prioritized key performance indicators is proposed, to be
followed by a phased and carefully managed upgrade down the road informed by successes, challenges,
and learning from piloting. The monitoring regime adopted by the national water sector regulator,
WASREB, provides a relevant benchmark for adaption by Kitui and other counties that the database
initiative would need to align to.
The database would underpin future decision-making by county governments and their stakeholders on
water services delivery, including planning and targeting of investments especially for O&M of already-
audited sources and development of new water sources. Such networking in a unified monitoring system
may further promote collaborative behavior among stakeholders through information sharing and
possibly lead to more impactful interventions. For Kitui County, the county-led WASH forum meetings
held quarterly with stakeholders and co-facilitated by UNICEF Kenya and SWS are an opportunity to
conceptualize and operationalize the database, continually examine performance of the water services
sector, and reflect on implications in terms of what needs to change to bring about sustainability.
62
Conclusion The audit results suggest that investments in county water infrastructure have grown considerably;
however, sustainability of water services remains a challenge. High non-functionality due to
infrastructure breakdown is a principal weakness, with far-reaching implications on reliability and other
water service quality indicators.
Kitui County falls within the arid/semi-arid climate zone, and an analysis of seasonal rainfall anomalies
reveals high intra-county variability, with the implication that selection of location and type of future
infrastructure investments will need to consider the inherent risk of drought, especially for surface
water sources. For instance, earth dam sources account for the second largest investment by value but
demonstrate a high vulnerability to droughts. High salinity was recorded and appears more pronounced
in some parts of the county than others, and further studies are necessary to develop detailed water
quality maps. Incidences associated with possible contamination of sources were also reported, yet 95
percent of the piped schemes do not treat drinking water before distribution. WMC’s capacity to collect
and use operational data is limited and thus there is an opportunity for ongoing capacity building to
improve financial management, accountability, and wider governance structures. The audit findings point
to the need for continued monitoring and targeted support of the rural water sub-sector for sustainable
drinking water services to be achieved.
The audit has provided a baseline status of water services as of December 2017. Meanwhile, many
changes continue to take place. Building on the data set by establishing a monitoring system to keep this
information up to date may be the most cost-efficient way for Kitui County Government to monitor and
manage its water services sector. This would involve development of a modest database in the early
phase to ease recognition of gaps and moderate dialogue on how best to progressively fill them, with
gradual improvements informed by learning. The database would provide a basis for Kitui County
Government to coordinate all WASH actors, monitor services, and ensure that their interventions
respond to emerging evidence from sector performance reports. The reports would also inform county-
and national-level strategies, polices, and targeting of investments.
The responsibility for universal water service delivery in Kenya ultimately rests on county governments,
a formidable task requiring major financial investments. There are prospects for county governments to
leverage the private sector for finances, technology, and expertise to complement their resources.
Nationally, the Government of Kenya through the National Treasury is supportive of PPP initiatives,
having enacted the PPP Act (2013) and Regulations (2014) that established a PPP unit to provide
technical, legal, and financial support to various government agencies, including county governments, to
undertake PPP projects. In addition, the Water Act (2016) allows county governments in Kenya to
engage the private sector for water services provision through contracting, and WASREB has, through
the collaborative efforts of national WASH actors, developed alternative models for water services
delivery in rural areas. One of the promising models is FundiFix, an insurance-based PPP rural water
infrastructure maintenance that has successfully worked in Kitui County and demonstrated results.
Therefore, creating better policies and incentives for the private sector, and investing resources to
accelerate the uptake of novel models of rural water service delivery, may reinforce efforts by county
governments to deliver a sustainable drinking water service for all.
A Water Infrastructure Audit of Kitui County 63
Annex A: Kitui County Water Audit Methodology
Objectives
1. To map and audit all of the water sources in Kitui County, including equipped sources (piped
schemes and hand pumps) and non-equipped sources (boreholes, rock catchments, shallow
wells, earth dams, springs, and sand dams).
2. To collect information in adequate detail to provide:
o an inventory of rural water infrastructure components;
o baseline key performance indicators and benchmarking related to technical operations,
financial management, and governance at water supply schemes; and
o a broader understanding of existing maintenance practices and challenges.
3. To provide key information for MAWL to assist with strategic planning, investments, and service
delivery improvements.
4. To provide key information for further development of models for water infrastructure
maintenance.
Outputs
The following outputs will be produced and shared with MAWL:
1. a summary report on the county water services situation; and
2. data files on all water sources mapped and audited (Excel).
Methodology
Collaboration: The activity shall actively engage MAWL officers (county and sub-county level) and
county administration officers (at sub-county, ward, and village levels). Meetings will be pre-arranged
with the management structure of each equipped source through the village administrators and sub-
county water office. Current management (committees, school boards, and individuals) will be asked to
prepare all their available scheme documentation for the meeting to increase the likelihood of
committee members and documents being available.
For non-equipped sources, no meetings will be pre-arranged, and information will be collected from
secondary sources including on-site locals or users, village administrators, signs, and other branding
installed on site. This is in addition to geodata and imagery that will be collected by the enumerator.
Respondents: Enumerators will conduct the survey with the management team (all adults over 18
years of age), who are responsible for operation and management of the water scheme. In the case of a
committee, the chairperson, secretary, and treasurer should be present where possible. Other
committee members may be present depending on availability.
Officials such as ward and village administrators will provide a preliminary list (name, type, contact, and
history) of all known water sources in their jurisdiction area but will be discouraged from being present
during the survey to avoid bias. However, results may be triangulated using separate interviews with key
informants where necessary.
64
Confidentiality and participation: Enumerators will introduce themselves, present the objectives of
the research, and explain that the information given will be available to MAWL, as well as used for
research purposes, and that participation is voluntary. All participants will be asked to confirm their
name, position, contacts, age (>18 years), and willingness to participate. Respondents can choose to
participate anonymously.
Data collection: A digital survey tool using the software ONA will facilitate data collection and analysis
using tablets. Key information on water source, infrastructure, operations, usage, water quality,
management, and finance will be gathered from participants. Some discussion and probing may be
needed to elicit the response, and answers will be cross-checked by the enumerator for consistency.
Additional details, observations, and qualitative data will be recorded by enumerators in a notebook and
submitted as “notes” in digital form. A mapping exercise will generate a sketch map of the scheme
components and nearby alternative sources, as well as key points of interest (schools, markets, clinics,
etc.), and photos will be taken of the different infrastructure components.
Where available, and where consent is given, photos of the following documentation will be taken:
• meter readings for the past 12 months,
• financial records (income and expenditure) for the past 12 months and bank statements,
• registration certificates of management body,
• constitution or by-laws, and
• minutes of the most recent management meeting.
Water will be tested for pH, EC, TDS, and temperature parameters using handheld Hanna meters.
Staff: Enumerators recruited will be required to:
• have had formal diploma or degree training in water resources technology, water engineering,
or other related field with experience in rural water supply;
• have some experience in a community engagement or facilitation role;
• speak fluent Kamba and be familiar with Kitui County; and
• attend a 5-day training on the survey tool, including a piloting exercise.
Duration
Approximate field time to be allotted to complete the work in seven sub-counties of Kitui based on
available data is 42 days, working with 14 enumerators.
Number of schemes: The audit will cover the more than 1,348 known water sources, as per the Kitui
County Government’s inventory.
To learn more about the Sustainable WASH Systems Learning Partnership, visit: www.globalwaters.org/SWS
University of Oxfordhttp://www.ox.ac.uk/Cliff Nyaga [email protected]