Safe Water Network

With mixed emotions, Charles Nimako put down the 2011 summary report on Safe Water Network’s projects in Ghana. His team, along with Safe Water Network leaders in New York, had worked for almost two years to make the Dzemeni (pronounced JEM-uh-nee) site profitable and provide safe drinking water to local Ghanaians. They had brought innovative ideas and actions to address enormous problems that had plagued water providers for decades; how- ever, difficult challenges and choices remained just ahead. He considered the two questions he and Hew Crooks, the organization’s SVP of Operations, discussed during their last phone call. How can we optimally expand the system to increase profitability at the Dzemeni site? And, should we move ahead with plans to expand the Dzemeni site into a regional “micro-utility” to provide safe water to the people living in nearby Tongor?
A native Ghanaian, Nimako had earned his MBA from Stanford University, worked as a consultant with McKinsey & Company in South Africa, and had most recently served as the CEO of the PepsiCo Bottler in Ghana and franchise director for East Africa.1 His work in the corporate world had proved invaluable in solving the many project problems encountered at the Dzemeni site, but the remaining challenges looked daunting.
Founded by American actor and philanthropist Paul Newman and a group of business and civic leaders, Safe Water Network sought to “develop innovative solutions that provide safe, affordable water to those in need.”2 Safe Water Network’s core values, as determined by the founders, are outlined in Exhibit 1. The organization brought together a group of dedicated professionals who had made their mark in private industry and civic work. For example, the founding CEO of Safe Water Network, Kurt Soderlund, also cofounded North Star Partners, a marketing and strategy consulting firm, and he had served as a special assistant to the president at New York City’s New School, assisting the president in his role as chairman of Human Rights Watch.3
Safe Water Network’s partner list features titans of industry from a variety of industries. The organization counts the Conrad N. Hilton Foundation, Kosmos Energy, the Merck Company Foundation, Newman’s Own Foundation, PepsiCo Foundation, Navajbai Ratan Tata Trust, and Starr International Foundation among its funders, and it has partnered with the International Finance Corporation (IFC), IBM, and the Johns Hopkins Bloomberg School of Public Health (JHBSPH). The Safe Water Network team on the ground in Ghana has contributed unique and valuable experience as well. In addition to Charles Nimako, the organization also hired Charles Yeboah to be its monitoring and evaluation, health, and hygiene coordinator, as Yeboah brought a public sector perspective from his work with the Ghana Health Service’s National Buruli Ulcer Control Program.4
Safe Water Network saw its role as a “Johnny Appleseed” of clean water, helping com- munities to jump start the journey toward clean water, relinquishing ownership and control, and then moving on to other communities to provide benefits. Safe Water Network currently operates in two countries: India and Ghana. The organization began in 2008 with the launch of a rainwater harvesting program in Rajasthan, India. The Indian operations expanded in 2010 with a 20-village safe water station initiative and the opening of an office in New Delhi. In 2011,
C-156

EXHIBIT 1 Safe Water Network Core Values20
Access
Making water available and affordable for all
Empowerment
Providing communities the confidence and competency for self-sufficiency
Impact
Realizing lasting health, social, and economic benefits
Measurement
Documenting success and failure
Lessons Learned
Adopting best practice
Environment
Safeguarding water resources
Risk Taking
Investing in new approaches
Open Source
Sharing our findings with the water sector at large

Safe Water Network provided technical assistance to the NGO Shining Hope for Communities in Nairobi, Kenya, in establishing a water tower in the Kibera slum to serve a girls’ school and the surrounding community. In 2012, they completed a market and feasibility assessment in collaboration with IFC to identify opportunities for commercial approaches to provide safe water to the poor in Kenya.5
The Ghana project started in 2009 when Safe Water Network funded the installation of five safe water stations in the country. WaterHealth Ghana (WHG), a local affiliate of WaterHealth International, approached Safe Water Network about an alliance: Safe Water would fund the sites and WaterHealth Ghana would manage and operate them. The physical and chemical process of water purification occurs at these safe water stations, as does the purchase of clean water by local consumers. Safe Water Network’s role in the project expanded after the initiative’s performance fell short of the business objectives of delivering safe water to thousands of rural and suburban Ghanaians, creating a sustainable and replicable model for commercial clean affordable water, and facilitating local ownership of the safe water stations. Safe Water Network partnered that same year with Johns Hopkins University, which conducted an independent study to evaluate the impact of the safe water stations on health outcomes.

The Dzemeni Project
Located on the Gulf of Guinea just north of the equator, Ghana gained its independence from Britain in 1957. Almost half of the country’s 24.3 million residents live in the coastal urban centers of Accra, Takoradi, the Gold Coast, and inland Kumasi; the other half reside in the country’s rural areas. After years of military dictatorships and coups, Ghanaians now enjoy a stable democratic government and a solidly growing economy. GDP growth has averaged a little more than 8 percent for the five-year period of 2007–2011.6 In spite of these impressive gains, Ghana remains a fairly poor country. Although the per capita income averages $1,230 annually for the country as a whole, 39 percent of rural Ghanaians earn less than $456 per year, or $1.25 per day, the global standard for extreme poverty.7

World Bank data indicate that 80 percent of rural Ghanaians have access to “improved” water; however, water systems fail at an alarming rate. For example, the Ghanaian Ministry of Water Resources, Works, and Housing found that 30 percent of hand pumps did not function at all and another 50 percent operated incompletely. Between 40 and 45 percent of rural Ghanaians, roughly 4.6 million people, lack consistent access to clean water. Money allocated by the government toward clean water projects often gets diverted to other uses; in 2010, more than 90 percent of funds budgeted failed to be spent. Water-borne diseases such as diarrhea, Buruli ulcer, and intestinal worms run rampant among the population, with 70 percent of diseases in the country traceable to the lack of safe water.8
Four of the five safe water stations served residents of the greater Accra peri-urban region (i.e., the area between the suburbs and the countryside), which include Amasaman, Pokuase, Obeyeyie, and Oduman. Exhibit 2 provides country-level data for Ghana, the two regions with safe water stations, and the five site locations. Dzemeni, located on the southeast shores of Lake Volta, represented the first attempt to reach rural and semirural Ghanaians. Lake Volta, formed in 1964 when the government dammed the Volta River at Akosombo Gorge, is one of the largest manmade lakes in the world.9 Dzemeni residents can draw water directly from the lake; however, decades of human and animal use have resulted in levels of microbial pollution that leave the water unsuitable for human use.
With about 7,000 residents, Dzemeni exists primarily as a market town and urban center, and it serves as a trading center for about 15,000 Ghanaians in the outlying hamlets. The Dzemeni safe water site made sense to Safe Water Network for several reasons. First, Dzemeni had no municipal water source and no other commercial water vendors existed; the convenience of Lake Volta effectively eliminated competition. Since it first appeared, residents have drawn water out of the lake, in spite of its contaminated state. Second, Safe Water Network

EXHIBIT 2 Country-level Data for Ghana
District Ga West South Dayi
Region Greater Accra Volta
Population 262,742 46,661
Population density (per sq. km) 370 46
Water coverage 19% 69%
% of communities paying tariffs 90% 60%
Water tariffs 10p/20L 10p/20L
Average size of population in LMS-ready communities 1,269 Nil
Average size of population in MSSF-ready communities 1,759 487
% of selected communities with electricity (LMS) 95% Nil
Potential population to be served 60,000 1,000

District Community Population
Ga West Amasaman 6,000
Pokuase 16,000
Obeyeyie 2,500
Oduman 2,500
South Dayi Dzemeni 7,700
Total 34,700
Source: SWN internal planning documents.

research indicated that more than 50 percent of residents, while poor, could reasonably afford to purchase clean water. Third, Safe Water Network was able to identify community leaders who would create excitement about safe water stations and encourage educational efforts that stressed the importance of clean water. Finally, nearby Tongor, which was also a combination of a number of villages, provided another potential market if the safe water station could profitably scale its operation.10
Delays and problems, however, cropped up almost immediately. Locating, screening, and selecting sites took longer than expected. Negotiations with community leaders at selected sites took time, and the sites required different water purification systems to address specific challenges. The Dzemeni center opened 18 months behind schedule, and capital costs for construction of the site exceeded its budget by 80 percent. Operating performance fell short of expectations; sales volumes indicated penetration rates (the percent of the population using the safe water station) around 20 percent instead of the 75 percent target, and the facilities failed to even cover operating costs.
In late 2010, Safe Water Network replaced WHG as the controlling partner. Safe Water managed health and hygiene education, marketing and demand development, and overall management at the Dzemeni site (as well as the other four sites). WHG continued its role as the Dzemeni site technical operator.11 Nimako and his local team began the hard work required to solve the many problems that stymied success. Safe Water Network began to learn that even the first goal, providing affordable potable water to rural Ghanaians, proved elusive and difficult, even more difficult than they expected.

The Water Users—Customers
Although the need for water is universal to humans, the demand for safe water is dependent on a variety of factors, including price sensitivity, convenience, seasonal variation in demand, and consumer knowledge about the benefits of clean water.

Price Sensitivity
Before the management transition to Safe Water Network began, in order to stem operating losses, WHG had doubled the price of water from 0.05 Ghana cedis (GHS) or 5 pesewas (p)— about $0.035—to 0.10 GHS or $0.07 for 20 liters (20 L). An average person needs 7 L of safe water for drinking and cooking each day, according to the WHO, and a family of five would use about 40 L per day. Overall demand dropped 30 percent immediately after the price increase, and the reduction in demand from the poorest Ghanaians, those Safe Water Network most wanted to reach, likely exceeded that figure. Incomes for the 40 percent living below the $1.25 per day poverty line averaged $1.08 per day, or $5.40 for a family of five.12 These families would have to spend about 2 percent of their income on commercial water, and studies from the Johns Hopkins group indicated that “cost did not seem to be a barrier to use.”13 In spite of increased operating costs, Safe Water Network ruled out any further price increases that would put the stations’ water out of reach of the poorest consumers, committing instead to achieving sustain- ability through increasing consumption and reducing costs.

The Convenience Factor
Water is heavy! A 20 L container weighs nearly 45 pounds, and not only did the spending of hard-earned cash for the water dampen demand but so did the physical energy required to transport water from the station to the home. Safe Water Network mapped water purchasers and found usage to be highly dependent on distance. For example, in Pokuase, they found that 85 percent of those living within 100 m (just short of the length of a football field) purchased water, but purchasing fell to 55 percent for those living between 100 and 200 m from the site, and only 10 percent of those living beyond 200 m made the effort to buy clean water. Dzemeni exhibited a similar pattern, with use dropping substantially for customers who lived further than 300 m from the site. Exhibit 3 details the effect of convenience on water consumption.
Safe Water Network decided to pipe water to remote kiosks (sales stations) some 400 m away from the main station to increase demand and consumption. The remote kiosks almost doubled sales at Dzemeni. The increased water sales, combined with the low incremental operating costs of the remote kiosks, significantly improved the economics of the safe water stations in these sites. Exhibit 4 illustrates the economics at the Dzemeni site. Total system cost per liter declined by 33 percent in Dzemeni after the remote kiosks were installed. Safe Water Network leveraged its positive experience and installed remote kiosks in two additional com- munities and early results are promising.14

The Effect of Convenience on Water Purchases

A. Household Mapping in Pokuase

100%

75%

50%

25%

0%
<100% 100–200
Distance from Site (m)

200

Source: Safe Water Network, “Decentralized Safe Water Kiosks,” [Internal Document] (August 2012). Available at http://www.issuelab.org/resource/decentralized_safe_water_kiosks_working_toward_a_sustainable_model_in_ghana. Reproduced with permission of Safe Water Network.

B. Remote Kiosk Impact on Sales Volumes in Dzemeni and Pokuase

1,000,000

800,000

600,000

400,000

200,000

–

Main Site/Bulk Remote Kiosks

1,000,000

800,000

600,000

400,000

200,000

–

Main Site/Delivery Remote Kiosks

Source: Safe Water Network, “Decentralized Safe Water Kiosks,” [Internal Document] (August 2012), http://www
.issuelab.org/resource/decentralized_safe_water_kiosks_working_toward_a_sustainable_model_in_ghana.

Comparison of Actual and Pro Forma Monthly Results at the Dzemeni Site

Monthly Results
Base Site 2 Remote Kiosks (Current) 2 Remote Kiosks (Projected) 4 Remote Kiosks (Projected)
Monthly Volume 400,000 600,000 7,500,000 1,000,000
Utilization 57% 50% 63% 83%
Realized Price 0.08 0.08 0.08 0.08
Revenue 1,600 2,400 3,000 4,000
Total Direct Operating Cost 1,196 1,591 1,736 2,131
Profit/(Loss) Before WHG Fee 404 809 1,264 1,869
WHG Fee 1,150 1,150 1,150 1,150
Profit/(Loss) After WHG Fee (746) (341) 114 719
Source: Safe Water Network, internal company documents.

A water delivery service was initiated by Safe Water Network to cater to convenience at the Accra sites. The price charged for truck-delivered water is 0.20 GHS for 20 L, twice the price charged at the main site. Costs increased as the service required a truck, driver, fuel, insurance, and maintenance. Even at the higher price, delivery sales were sufficient for the service to cover its operating costs. The service would need to grow its volume by an additional 25 per- cent, however, if it hoped to cover its capital costs—the trucks and containers.15 Nimako and his team identified and explored several potential avenues for expanding the reach and purchase penetration at Dzemeni, including truck and donkey delivery, additional remote kiosks, and contracting with local shop owners to sell water. Delivery services did not exist in Dzemeni, and Nimako wondered about how such a new distribution method would affect demand.

Seasonality
As with rural populations around the world, Ghanaians have relied on rainwater harvesting to provide fresh, potable water during the rainy seasons, which in southern Ghana extend from April to July and from September to November. The annual rainfall average in Ghana is more than 80 inches per year, a foot more rain than in Mobile, Alabama, the rainiest location in the continental United States.16 The demand for commercial water could potentially vary 50 to 65 percent between the dry and rainy seasons, depending on the intensity of the rainy season. Demand for clean water would be strong and require sufficient capacity during the dry season, but Ghanaians would naturally economize during the rainy season and collect free water. Safe Water Network chose to install enough capacity to cover the dry seasons, and believed it could stimulate enough demand for commercial water during the rainy season.

Education and Recontamination
Many people in developing countries do not understand the need for consistent clean water to combat water-borne diseases, and this lack of understanding leads to mishandling and recontamination of clean water by consumers. A list of common myths, and the truths that refute them, can be found in Exhibit 5. Since the beginning of the initiative, Safe Water Network used local health and hygiene teams to engage in community education and outreach. Dzemeni’s cultural tradition of strong local leaders (chiefs) meant that, after Safe Water Network taught them, others in the community would likely become informed. Safe Water Network did not rely solely on the chiefs, and the organization recruited other influential volunteers, including local women, teachers, and even primary school children (peer educators) to teach residents about the need for, and benefits of, clean water.
The downside of this intervention came at a cost—$200 (285 GHS) per month added to the Dzemeni operating costs. The upside could be sustained with increases in demand for commercial water. However, an internal report noted “the full effects of health and hygiene education and resulting changes in behavior may not be realized for some years.”17
The open or multiple-use water containers that people used, and lack of hygienic practices, often resulted in recontamination of the treated water. Consumers dipped their hands into their clean water and introduced contaminants to the water in the container, and those contaminants remained in the container to pollute the next batch of clean water purchased. To solve this problem, Safe Water Network introduced narrow-mouthed containers because the smaller opening kept hands out. Solving this problem created another, however: the cost of the containers. Two thousand narrow-mouthed containers were sold to Dzemeni residents along with a coupon that provided users with vouchers good for 30 containers of free water if they purchased a container for GHS 3 ($2.10). Although most consumers would willingly accept a subsidized container, it was not clear what the demand would be if they had to purchase them at the full cost.
Charles Nimako recognized that Safe Water Network had learned much about how people use—and whether they will buy—water. The result of their efforts was a 151 percent increase in overall demand from late 2010 to late 2011. That impressive figure encouraged both Nimako and

EXHIBIT 5 Consumer Education Materials
Overturning Myths, Selecting Messages, and Identifying Influence Leaders to Promote the Safe Water Station
Section 1: Myths That Need to Be Addressed in Promoting the Safe Water Station
MYTH #1: Groundwater is pure.
TRUTH: Groundwater can contain many kinds of chemical and microbiological contaminants that are natural or generated by humans. These can cause diarrhea, cholera, and arsenic or other chemical poisoning.
MYTH #2: Water cannot cause illness since illnesses are sporadic and not constant, whereas water is drunk daily.
TRUTH: Illnesses can be present all the time but only be apparent when the body is overpowered by the contaminant causing the illness. Dirty water may not always cause illness, but safe water and hygienic practices will always prevent it.
MYTH #3: Groundwater and water front natural streams are pure.
TRUTH: Both water sources can be contaminated by people, animals, or naturally occurring pollutants. Our water quality tests can show this.
MYTH #4: Boiling treats the water of its contaminants.
TRUTH: Boiling water can eliminate microbiological contaminants and help reduce diarrhea. It is, however, useless against chemical contaminants including arsenic, fluoride, and “chalk” or hardness.
MYTH #5: Consuming safe water sporadically can prevent health issues.
TRUTH: Safe water needs to be consumed by all family members at all times to prevent health problems. Drinking unsafe water occasionally can contaminate both individuals within the household and the places where it is stored. Occasional use of unsafe water can make safe water unsafe!
MYTH #6: Schemes are not to be trusted.
TRUTH: We know this is often the case. But the Safe Water Station is owned and managed by friends and neighbors in your own community. You can contact them or the other Safe Water Station Partners at any time if you require any information that will encourage your trust. Also, the water that is sold must pass strict testing requirements that ensure the quality of the Safe Water Station water is as good as any sold anywhere in the country—and they are required to make the results of these tests available to customers.
Source: Safe Water Network, “Tools for Safe Water Stations,” p. 125.

Hew Crooks, the SVP of Operations, but it did not solve their fundamental problem. Even with the impressive increases in demand for Safe Water products, the revenues generated covered only the operating expenses. Nimako wondered whether demand would ever cover the full costs of the Safe Water Stations and whether the fundamental business model and cost structure of Safe Water Network would need to change in order for the company to stay in business.

The Safe Water Network Business Model
Safe Water Network envisioned solving the challenge of potable water by creating financially sus- tainable and community-welcoming safe water facilities. The organization believed that if it could accomplish these goals, a third goal could be reached: finding local owners to buy and run the safe water stations. Taken together, these three overarching goals constituted the Safe Water Network objective. Nimako had an emerging understanding of the consumer side of the safe water initia- tive, and he mentally reviewed each pillar of Safe Water Network’s business model. Exhibit 6 shows the Safe Water Network process for implementing the safe water initiative in a rural community.

Financial Sustainability
Rain water is pure; the evaporation process removes inorganic pollutants—chemicals, sedi- ments, residues—as well as bacteria or other organic contaminants. Unfortunately, many “fresh water” sources in Ghana, such as Lake Volta, streams, wells, or springs, provide con- taminated or impure water because of the presence of inorganic chemicals and minerals or

The Safe Water Network Implementation Process

Source: Safe Water Network, Tools for Safe Water. Stations, p. 125. Reproduced with permission of Safe Water Network.

microbial contaminants associated with human or livestock fecal matter in the water. The essence of Safe Water Network’s mission involves purifying impure water by moving it through a treatment system. The physical and biological characteristics of the water source drive the choice of treatment technology, which in turn underlies the financial sustainability of the Safe Water Network business model.
The technology then determines the capital cost of the system. The cost of equipment identified as suitable for community-sized systems in Ghana ranges from $23,000 (32,600 GHS) to $100,000 (142,000 GHS), not including land, building, and permitting costs (see technology descriptions below). Consumables such as sand, gravel, chemicals, membranes, operator and staff salaries, and maintenance expenses constitute operating costs. Safe Water Network has no allegiance to any particular technology, and leaders seek to install the most appropriate and cost-effective technology for a given site. Exhibit 7 describes the advantages and disadvantages of each treatment technology.
Carbon filter with ultraviolet light treatment (CUV). All five initial sites in Ghana used CUV- based systems. These systems employ sand, carbon, and 10- and 1-micron polypropylene wound cartridge filters in addition to ultraviolet (UV) treatment systems. Water passes through a carbon filter and a series of other filters to remove chlorine, sediment, and other inorganic compounds. Exposure to UV light alters organic DNA and kills bacteria, viruses, yeast, mold spores, fungi, algae, and fecal coliform.18 CUV systems cost $50,000–$90,000 to purchase and install. This system may prove to be the most costly technology to operate because it requires well-trained technicians to operate the systems.
Limited mechanization systems (LMS). LMS involves installing a submersible pump in a newly drilled or existing borehole (well), pumping water to an elevated water storage tank, and distributing the water via gravity to standpipes. This system does not provide water filtration; it only provides more water from existing boreholes. The water produced must be chlorinated before distribution to guarantee purity. The cost of LMS varies according to number of bore- holes and pumps and the size of the storage tank. Safe Water Network’s budgeted LMS costs are $23,000 for 1,000 people, $32,500 for 2,000 people, or $48,000 for 5,000 people. The use of solar power would reduce operating costs but increase capital outlay by $5,000–$6,000. Safe Water Network is evaluating the use of LMS technology for a possible expansion in the Eastern Region.
Modular slow sand filtration (MSSF). Slow sand filtration (SSF) has been in use since the early nineteenth century; however, in the last few years, the design and construction of modular systems has made MSSF an attractive method for smaller communities in the developing world. The system is appropriate and suitable for perennial surface water sources with low turbidity (mud and sediment). The system consists of twin modular plastic tanks for use as the main filter, supported with plastic raw and treated water storage tanks. The treated water is distributed by gravity, pumping, or a combination of the two, depending on the size of the system and the location and distribution of a remote kiosk. MSSF systems run $43,000 for 1,000 people,
$47,000 for 2,000 people, and $69,500 for 5,000 people. Safe Water Network is employing MSSF at two new sites launched in the spring of 2013, in preparation for a larger-scale MSSF-based expansion planned around Lake Volta.
Ultra-filtration (UF). UF utilizes a range of polymeric-based semipermeable membranes to filter particulates of different sizes. UF provides a filtering solution for communities with sur- face water sources but variable levels of turbidity or contamination. Primarily used in industrial

EXHIBIT 7 Treatment Technology Advantages and Disadvantages
Technology How It Works Pros Cons
LMS (Limited Mechanization Systems) Mechanizes existing bore- holes; can use solar power • Can leverage existing 24,000 boreholes in Ghana; low initial outlay cost
• Low operating costs; costs of purchasing solar energy materials are dropping and are expected to continue to drop
• 77 promising sites in Ghana for LMS technology
• Builds on the investments of existing bore- holes at a very low incremental unit cost to significantly improve convenience • Requires a high-yield borehole or well
• Requires pure water source supplemented by chlorine
• Requires soil and water tables that allow for bore holes—not useful near large bodies of surface water with a shallow water table or where soil is contaminated
MSSF (Modular Slow Sand Filtration) Addresses microbial contamination in surface water; can use solar power • Flexibility to expand with demand, avoiding overspending on unneeded capacity; relatively easy to operate
• Minimal maintenance costs
• Costs of purchasing solar energy materials are dropping and are expected to continue to drop • Requires perennial surface water source with low turbidity
• New approach in Ghana (untested)
• Effective for microbial contaminants for the most part but more aggressive para- sites like viruses may not be filtered
Membrane Ultra-Filtration (UF) Appropriate for communities whose water needs aren’t served by LMS and MSSF (have highly variable turbidity/contamination) • Able to treat surface water with high or variable turbidity, certain metals (including iron and manganese, which require pretreatment), color bodies, and a broader range of microbes (including viruses) may not be reduced as efficiently in SSF systems
• Treated water will be of adequate quality regardless of operational challenges
• UF units are modular and can be scaled as demand changes • Higher outlay costs; need a population of at least 3,000 to achieve lower cost per capita water production
• Because of high costs, considered an option only when conditions require this technology (this technology is superior and has fewer cons other than the fact that it is very expensive and difficult to fit a viable business model)
Carbon Filter w/Ultraviolet Light Treatment • Known and tested technology
• The only viable turnkey solution for treating microbial contaminated surface water when the Dzemeni project was initiated
• Multiple technologies can be combined depending on source water • Requires skilled technicians to operate.
• If filters or UV bulbs are not maintained/ replaced, system still operates but pro- duces contaminated water
• High initial capital expense
• Not easily scalable
Source: Adapted from Safe Water Network internal planning materials.

settings, the technology has recently shown potential for adaption in developing world set- tings. Its wide range of operational modes (pressure, gravity, chemical addition) allows for broad applications in a variety of settings and input water qualities. UF systems range between $50,000 and $100,000; capacity is not an issue (a UF station could filter 2,000 L per hour), but costs vary based on the filtration method used. Safe Water Network is using UF for two new sites in the Western Region.
The availability of electrical power also plays a significant role in technology and cost. Access to the electrical grid represents one challenge of working in rural Ghana, because the provision of safe water must depend on the reliability of another system. Solar power provides a possible alternative, but the Safe Water Network will incur greater capital costs in exchange for lower operating costs. A major benefit of solar power is the possibility of creating remote charging stations for appliances, cellphones, and batteries. If the solar recharging sites generate 100 to 200 GHS ($70–$140) per month, it would contribute to the overall profitability of the safe water station. Safe Water Network launched its first sites using solar power in 2013.

Welcoming Communities
Several criteria factor into the decision to build a safe water station in a community. Safe Water Network hoped to locate in communities with greater than 4,000 people or 800 families within reasonable proximity to the station to provide adequate demand for clean water. The type, nature, and prevalence of water-borne diseases in the community affect the type of treatment system and the overall value added from clean water. Finally, Safe Water Network had learned through its India initiatives that communities without a commercial water market increase the level of difficulty—and time required—to get people to adopt clean water. Again, Exhibit 5 dis- plays common myths that often stymie adoption within a community. Asking residents to pay for treated water when they were accustomed to collecting untreated water for free added a new layer of resistance to Safe Water Network initiatives.

Local Ownership
The exit strategy for a safe water intervention in a community calls for local, community-based groups or individuals to take control of the station after capital costs have been recouped by Safe Water Network. Local ownership could dramatically improve adoption rates, assist in consumer education, and provide employment and income for a community. Safe Water Network saw its role as helping communities to jump start the journey toward clean water, relinquishing ownership and control and then moving on to other communities to provide benefits. Safe Water Network’s strategic planning materials identified several characteristics of ideal local owner/partners. These included credibility with local stakeholders, competence in running a business and fulfilling the duties of a good partner, and high ethical and moral character.
The costs for the Dzemeni site illustrate the revenue model for all Safe Water Network safe water stations. Capital costs to install the UV filtration unit to service the almost 8,000 residents in the area totaled $91,000, almost twice the budgeted amount. Operating costs include staff and technician labor, consumables such as cartridge filters, UV light sources, electricity costs, maintenance, and repair. The pro forma costs displayed in Exhibit 4 detail the basic economics of running the Dzemeni site, including a management fee payable to WHG.
Nimako saw his team’s hard work reflected in the operating profits at Dzemeni. After nearly two years, the Dzemeni site generated a positive cash flow from direct operations; unfortunately, the management fee, negotiated at the beginning of the project, paid to WHG turned profitability negative. Reducing the fee represented one hurdle to financial sustainability. The Safe Water Network team believed that, if a cluster of safe water stations and accompanying kiosks could be installed close to the Dzemeni site, they would be able to amortize the flat management fee over a number of independent profit centers. Exhibit 8 presents the expected capital budget for a new safe water station in the neighboring community of Tongor.
The Tongor site would service another six surrounding communities and about 11,000 people, and the site would allocate the management fee among a number of communities that the team calls micro-utilities.19 These utilities might create other economies of scale in consumer education, demand generation, and other operating activities such as repair and maintenance.
Safe Water Network’s initial business model sought not only operating profitability and the creation of a small reserve, but also the recovery of the station’s capital costs. The organization viewed its capital investments as loans to the facility with a payback period of eight years. With an imputed interest rate of 2 percent, the Dzemeni site would fail to pay back its $90,000 (135,000 GHS) capital loan even if the two remote kiosks could generate almost ($500,000) 750,000 GHS in total sales. If the organization could create a micro-utility to service a number of communities, capital costs could be recovered in the appropriate time horizon.
The underlying appeal of local ownership came from the potential of the safe water stations to provide income for a local partner/owner. The Safe Water Network financial models predicted a monthly profit for owners or communities of $90–$100 per month, just about the per capita income for Ghana. To date, the Dzemeni project team had invested little time and energy in identifying and selecting potential long-term partners. The team’s time had been

EXHIBIT 8 Projected Capital Costs at Tongor (CUV Treatment Technology)
Uses US$ GHS
Local civil works (small office block, base of tanks, trenching, etc.) $ 20,000 30,000
Steel tank (Blue Filter quote) 150m3/day capacity 31,000 465,000
Tank cover 6,000 9,000
Shipping and clearing 8,000 12,000
Tank welding supervisor (per Blue Filter quote) 10,000 15,000
Blue Filter supervisor 5,000 75,000
Raw water storage (3 Polytanks of 25k litres each) 12,500 18,750
Treated water storage (3 tanks of 25k litres each) 12,500 18,750
Overhead tank 45,000 67,500
Local design and consultancy 20,000 30,000
Chlorination system 2,000 3,000
Electrical connection to source (including transformer) 6,500 9,750
Service points (6 & 3k) 18,000 27,000
3hp submersible pump 8,500 12,750
4 inch HOPE pipes (from source to plant) 15,000 22,500
2 inch HOPE pipes (for supply of treated water) 25,000 37,500
Sand media 5,000 7,500
Source pump sump 5,000 7,500
Working capital (post-opening) 15,000 22,500
Contingency 10,000 15,000
Total Uses 280,000 420,000
Source: Safe Water Network, “Tongor-Dzemeni Case Study” [Internal Document] (June 2011).

filled with learning the business in the communities and achieving operational profitability. No one knew how easy it would be to find a partner with the passion, skill, and vision to run a safe water station even if the financial rewards were in place.

Conclusion
Nimako and his team had made significant progress at Dzemeni. The team had created demand for safe water, the site was now operationally viable, and the Ghana and New York staffs both saw in Dzemeni the seeds of a financially sustainable solution to the clean water problem. Now that the Dzemeni system had established a track record of successful operations over the course of three years, Nimako wondered how the organization could optimally expand the system to serve the surrounding communities.
What knowledge gained from their experiences should guide the development of a fully functional micro-utility system through both (1) expansion of safe water services into other nearby communities and (2) additional points of sale or distribution methods within Dzemeni. Nimako knew more sales points would improve access, convenience, and usage, but the existing pay-as-you-fetch model worked only for relatively centralized, high-volume remote kiosks. (Basically, the first 50 users paid the operator’s salary.)

Likewise, if the company pursued a multitown micro-utility, what elements of the process should it replicate in order to deal with the inevitable cultural differences, even among local villages? How would the company solve the management challenges? Private system operators in Ghana have said that they would not be interested in managing a system as small as Dzemeni, so how would Safe Water Network select and train entrepreneurs to run a system that spanned several communities?
Nimako began to think of some larger, more strategic questions as well. For example, should Safe Water Network partner and engage with the local or Ghanaian government? Government leaders often made promises to build municipal systems, usually during the election season. Municipal systems would represent a formidable competitor to the Safe Water Network’s model. Should Safe Water Network partner with the government to forestall entry and protect their investments? What opportunities existed to employ lower-cost treatment technologies than the system currently installed? The Dzemeni model worked for aggregated communities of 7,000 or more people, but could the organization create safe water stations to service smaller communities? Finally, should Safe Water Network rethink its financing approaches to support capital investment? Should it shift the business model to targeted philanthropic investment (with no repayment) rather than the current loan-based model? Would the financial rewards of ownership produce the outcomes Safe Water Network predicted and create dedicated and skilled local partners capable of running the community water business?

References
1http://www.safewaternetwork.org/team.asp, accessed March 8, 2013.
2Internal Safe Water Network, FAQ, https://www.safewaternetwork
.org/faqs, accessed February 19, 2013.
3K. Soderland, http://www.safewaternetwork.org/Team.asp, accessed February 20, 2013.
4Information on the leadership team of Safe Water Network taken from http://www.safewaternetwork.org/team.asp, accessed February 4, 2014.
5Information drawn from internal Safe Water Network documents.
6Calculated from data available through the World Bank, http://databank
.worldbank.org/ddp/home.do?Step=3&id=4, accessed February 20, 2013.
7Ibid.
8Safe Water Network, “Decentralized Safe Water Kiosks” [Internal Doc- ument] (August 2012), p. 3.
9Safe Water Network, “Tongor-Dzemeni Case Study,” [Internal Docu- ment] (June 2011), p. 5.
10Ibid.
11Ibid.
12This figure was calculated from the Poverty Gap Index, a measure designed to assess the intensity of poverty for those living in extreme poverty. The gap measures the average measured gap between the actual measured household income and the poverty line. For Ghana, the

rural poverty gap is 13.5 percent, meaning that the average household income of those living below the poverty line is 13.5 percent below
$1.25 per day. Poverty gap data found at http://www.tradingeconomics
.com/ghana/poverty-gap-at-rural-poverty-line-percent-wb-data.html, accessed February 19, 2013.
13Safe Water Network, internal documents.
14Safe Water Network, “Remote Kiosks: A Path to Improved Economics and Coverage” Internal Report.
15Decentralized Safe Water Kiosks, pp. 8–9.
16Data on Ghana taken from http://www.ghanaweb.com/GhanaHome Page/geography/climate.php, and data for the United States taken from http://www.livescience.com/1558-study-reveals-top-10-wettest-cities
.html, accessed February 20, 2013.
17Decentralized Safe Water Kiosks, p. 10.
18A description of different treatment technologies can be found at Engineers Without Borders, http://www.stevens.edu/ewb/docs
/ME423_Fall_07_Phase_III_Group_10_EWB_-_Water_Purification.pdf, accessed February 21, 2013.
19Personal communication between the case writer and Safe Water Network leaders.
20http://www.safewaternetwork.org/Values.asp, accessed February 4, 2013.