1 7

2 Estimating costs of small scale

3 water supply interventions
4 Paul Jagals, Luuk Rietveld
5 INTRODUCTION
6 Within the scope of this Guide, the main objective of this chapter is to explain a
7 basic approach to estimating financial costs of installing, maintaining and
8 operating a small scale drinking water supply. After assigning shadow prices to
9 certain costs as appropriate, the outcome of such an estimate can then be used,
10 with the estimates of the total benefits derived from water improvements, to
11 select the best improvement or intervention for a given target group in the
12 context of their livelihood patterns, by comparing rates of return.
13 Costing is one step in the economic assessment of water supply projects and
14 an economic assessment is one part of the full set of information (economic,
15 environmental, health, social, technical assessments and feasibility studies)
16 likely to be used by decision makers in selecting the specifications of the system
17 to implement. In the context of this chapter, the costing method proposed is not
18 intended to provide guidance on full economic valuation of the costs, but rather
19 to provide a financial input into a cost-effectiveness or social cost-benefit
20 analysis of small scale water supply improvements. In a broader sense, costing
21 is an essential element of any economic analysis that may involve modifying
22 financial costs through the assignment of shadow prices to reflect true economic
23 value.
24 At this point in the process, the aim is to identify the financial costs. The
25 description of the costing method in this chapter therefore aims at providing an
26 incremental price in present-day monetary terms (year zero) of water supply
27 technology providing water to a community against which the derived benefits
28 could be measured, so that informed decisions can be made. Information on full
29 economic costing can be found in the WHO publication, Costing Improved
30 Water Supply Systems for Low-income Communities, A Practical Manual,
31 developed by the Department of Econometrics (DE), Faculty of Economic and
32 Social Sciences, University of Geneva (UNIGE) (Carlevaro and Gonzalez
33 2008).
34 The present Guide is describes the method of social cost-benefit analysis for
35 use at a national level (though sensitive to local livelihood patterns), with the
36 intention that it can be used by non-specialists and specialists alike. So to keep
37 the presentation accessible to all readers, costing of items that are incurred for
38 activities at national level or are indirect outcomes at national or local level are
39 not dealt with in this chapter. These are costs such as external environmental
40 costs, which arise out of local environmental damage or protection, opportunity
41 costs which value the forgone benefits of diverting raw water from productive
42 activities like agriculture (both with significant livelihood implications), to non-
43 productive ones like basic domestic uses, and depletion premiums which value
44 the loss of water supplies from difficult-to-replenish sources, or a share of
45 overhead costs that are needed to run national regulatory and laboratory
46 facilities.
47 The approach followed in this chapter to estimate the use costs of a water
48 supply intervention such as costs for construction, operation, maintenance,
49 direct administration and overheads. It provides an insight into simple and
50 effective costing that a specialist would use and which a non-specialist at
51 national as well as local level would understand. In practice, costing preferably
52 should be done at the service provider level e.g., a regional water management
53 body or at lower levels, a district authority. The costing could, depending on
54 local capabilities, also be done by the end-user target group or a local NGO.
55 Costing a system locally implies that a water service provider i.e. a district
56 authority, or a local user group, decides to invest in improved infrastructure (i.e.,
57 storage and treatment facilities and a distribution network). After constructing
58 and activating the system, the service provider then continues to spend money
59 on the system for operation, maintenance, future rehabilitation, administration,
60 training, promotion and education (e.g. energy and chemicals use), and has to
61 make sure that these expenses are covered by some form of income (standing
62 charges, consumption rates or subsidy). The main objective of the service
63 provider might not be to make a profit, but to provide an economically efficient
64 water service that can also provide benefits in improving the target group’s
65 livelihoods. The purpose of a cost estimate in this situation is therefore to assist
66 decision makers responsible for the provision of services with a reliable
67 financial value for providing improved drinking water supply technologies.
68 Chapter 6 describes the differences in quantities or quality of services that
69 characterize small water system technologies. A useful characteristic of small
70 systems is that they can allow for incremental improvement as the target group’s
71 needs change over time, perhaps as a result of population growth or move from
72 standpipe to in-house provision. This may lead to a decision to initially choose
73 the most affordable system (least-cost system) with the intention to
74 incrementally adapt the system’s capacity to fit the needs or financial
 75 capabilities of a growing population. A practical costing approach must allow
76 for this incremental costing as well.
77 This chapter discusses the methods and procedures for costing a small water
78 system in three sections. Firstly, the implications of installing a small water
79 system intervention are addressed, be it in the form of putting in a system where
80 there wasn't one before, or upgrading an existing system. Secondly, a brief
81 discussion follows on the various types of financial costs that will be
82 encountered in the process of costing a small water system, and lastly a
83 simplified costing approach is presented to reliably estimate a pattern of costs of
84 systems that allows for costing across time should a system be incrementally
85 adapted.

86 IMPLICATIONS OF COSTING A SMALL WATER

87 SYSTEM INTERVENTION
88 The local costing an intervention system may be done for a new system or for
89 upgrading an existing water supply system to a more sophisticated system.
90 Costing a new system for an area where there is none at first might be quite
91 complicated as it takes place in the early planning stages before the actual
92 construction begins to take shape. Especially costing of the initial stage of a
93 project to install a water system will potentially have large inaccuracies. One
94 reason for this is that relatively little is known during the initial stage
95 (configuration of the treatment scheme, specific construction demands, and local
96 conditions). Another reason is that many things can and will change during
97 subsequent designs. This all requires great understanding and integration of
98 three critical aspects i.e. the service provider must have a good idea of what type
99 of intervention (i.e. system) might potentially be installed e.g. the intervention
100 types in Chapter 6 will provide a good idea.
101 With this in mind, the layout of the system must be determined (relative
102 positions and elevations of source, storage and pipe network configuration) and
103 then the elements of the potential system must be sized. Only then can the
104 costing of the intended system commence. While costing can be done in great
105 detail to cover for these uncertainties, this will require a high level of
106 collaboration between service providers and engineering planners. It should
107 therefore preferably be done by such specialists.
108 For detailed information, including comprehensive checklists, reference is
109 made to the Practical Manual by Carlevaro and Gonzalez (2008).
110 Costing an upgrading of an existing system is more readily attempted by
111 specialists or non-specialists alike because of the fewer degrees of freedom (or
112 uncertainty).
113 After following these preceding steps described above, estimating the
114 approximate cost of a small water supply system can start, usually with
115 estimating the investment or capital costs. However, the service provider
116 should look at the complete picture and costing therefore has to be extended to
117 include recurrent (operation and maintenance) costs. This second component is
118 absolutely vital to predicting what the sustainable operation of the system is
119 going to cost the service provider once the system is built. If this is neglected,
120 the intervention will be short lived and the benefits often negated before they
121 were accrued. This will be discussed further in the next section.
122 Cost estimation necessarily requires a large number of inputs. In order to
123 simplify the data collection and preparation steps, a three-tier data structure is
124 required. The first data category captures the engineering parameters, which
125 would typically include technical specifications such as pump and motor
126 efficiencies, as well as pipe friction coefficients, which are not likely to vary
127 significantly anywhere in the world. Data for these parameters should only be
128 measured by those with an engineering background with the capacity to provide
129 a sound technical judgement based on experience. The second data category
130 captures the monetary parameters, which would typically include the cost of
131 pumps, pipes, holding tanks, fuel and electricity, as well as the ratio between the
132 costs of labour and materials for system construction. These parameters will be
133 fairly constant for any particular economic zone. Once these parameters have
134 been calibrated for a particular region, they can be left unchanged while
135 different water supply systems within the region are analysed. The third data
136 category captures the system parameters, which will typically include pipe
137 types, diameters and lengths, storage tank volumes and the number of
138 standpipes. These parameters are unique to each water supply system, and have
139 to be determined on site or from engineering drawings.

140 TYPES OF COSTING

141 Costs by definition consist of all resources required to put in place and maintain
142 the intervention. These include capital costs (investment in planning, preparing,
143 construction, purchase of hardware) and recurrent costs (operation, maintenance
144 and monitoring). The cost of a small water supply system usually includes
145 capital as well as recurrent costs in each of the usual activities of a water supply
146 i.e. source, treatment, pumping, storage and distribution. The costing method
147 must be robust and it will need to provide reliable estimates by aggregating
148 collections of physical parts of a water supply activity into single units of cost.
149 An example is the estimation of the initial costs of water treatment for a new /
150 improved system. To get started, cost-functions can be used that are based on
151 previously completed projects for which there must be several examples in any
152 country. Often in such functions, the design capacity is typically included as a
153 variable e.g. cost per m3 of treated water and this will give the planner a good
154 estimate of the cost of water treatment for a village of X number of people
155 consuming Y litres of water per person per day (ℓ/c/d).
156 While this example is simplistic and robust, the costing model at the end of
157 the chapter is more refined but is still based on this approach.

158 Capital costs

159 Formally defined, capital goods is the stock of goods which are man-made and
160 used in production (as opposed to consumtion). Fixed capital goods (durable
161 goods such as buildings and machinery) are usually distinguished from
162 circulating capital goods (stocks of raw materials and semi-finished goods
163 which are rapidly used up). Capital costs are the costs incurred by emplying
164 capital goods. In accounting conventions, capital goods are usually taken as
165 those with a life of more than one year, such as land, buildings and equipment.
166 In the context of developing and installing a small water system the capital
167 costs represent, therefore, the total costs that are not expected to recur for
168 significant periods of time. These are costs for the preparation and construction
169 of the system through to the moment that the system becomes operational.
170 From that moment on, the system must be operated and maintained towards the
171 optimal benefit expected from the intervention i.e. the O&M costs. Capital
172 costs can also occur during the operational lifetime of the system. Examples
173 include expansion of the system and replacement of major (high-cost) parts.
174 Capital costs usually include those costs related to the construction and
175 equipment activities of installing the new system. These costs flow from the
176 preliminary studies, which are conducted during the pre-investment (e.g.
177 planning) stage and involve the study of the technical, economical, social,
178 environmental and health aspects in the construction project.
179 A drinking water system consists of a variety of fixed (constructed)
180 installations like filter units, clear water reservoirs, pipes etc. Depending on the
181 size of the system, these construction activities might include office and sanitary
182 facilities for the staff of the new treatment facility and / or a workshop and
183 maybe a small laboratory with facilities for the maintenance personnel. Besides
184 these costs, the furnishing of staff facilities, workshop and a laboratory is part of
185 the capital costs. The project requires equipment, which will be a capital cost in
186 items such as pumps and power systems. Materials are needed to complete the
187 construction including materials bought or acquired by the community or the
188 municipality in the local markets of the country as well as imported materials.
189 The workforce for the construction can be specialists such as engineers,
190 constructors, technical staff, and social science professionals. It will also
191 include non-qualified workers that work in excavation, cleaning, etc. These can
192 be the people from the local community. Lastly, semi-qualified workers will be
193 required - generally a type of worker between qualified and non-qualified. This
194 will depend on the work activity.
195 Other capital costs will be related to management of the project,
196 administration, direction, coordination, logistic, transportation, communications,
197 office costs, private executors and control of quality, and any other unassigned
198 cost of the project. Contingency costs are an amount or percentage of the total
199 capital costs included in a project account to allow for adverse conditions that
200 will add to the basic costs.
201 A cost which will often be encountered that should be seen as part of the
202 capital costs will be the acquisition of land that might be required for
203 components of the system e.g. the site for the treatment facility, land covered by
204 water when a surface source such as a stream or river is impounded.
205 Provision must also be made for overheads and supervision. Once all the
206 capital investment costs have been estimated, their sum will reflect the net
207 construction cost. A contractor, be it the villagers themselves or outsiders,
208 might add a surcharge to allow for site establishment, site clearing, supervision,
209 profit, etc, which are all allowed for by adding an additional percentage to the
210 net construction cost. (A typical surcharge for contracts in rural South Africa is
211 25%.) This then add up to form the total contract cost. For a new water supply
212 system, the client has to also bear the costs of planning, surveying, soil
213 investigation, possibly exploratory drilling, contract management, quality
214 control, etc. These design and supervision costs, paid to consulting engineers or
215 borne by the client’s own design staff, amount to an additional surcharge (about
216 25% over and above the total contract value) to finally determine the total
217 project cost.

218 Recurrent costs

219 These costs comprise all expenditures (staff, parts and materials) that are
220 required to keep a system operational and in good condition (maintenance) after
221 its installation was completed, including certain fixed costs, which, depending
222 on the finance policy of the service provider, need to be provided for on an
223 annual basis. An example of this is the creation of a replacement fund through
224 annual depreciation levies. Monitoring of the system can be an operational
225 function or could be a regulatory function to ensure the quality of the water
226 supply to the community. It has a cost that can become part of the recurrent
227 costs as a separate item or part of O&M costs, depending on the need and the
228 extent of the system. This illustrates a very important point in the context of
229 costing. For small systems management it is often the case that maintenance
230 costs are budgeted for annually at the service provider level, which is usually a
231 tier above local community level. Operational costs are usually budgeted for at
232 the local level.
233 The maintenance costs contain all costs for the repair and replacement of
234 parts of installations (e.g. pumps or repairing wells) within the predicted lifetime
235 of the small water supply system, as far as these are not included in the
236 operational cost. Effective maintenance is the key to sustainability of a system
237 but is often neglected in terms of its effective execution, rendering many small
238 systems ineffective not long after their inception.
239 In general operational costs are considered to be mostly costs for acquiring
240 and administering consumables such as energy, process water and chemicals, as
241 well as disposing of waste. These consumables do not include general
242 maintenance materials such as paint, lubricating oil, and tools as these should
243 included in the maintenance costs.
244 Fixed costs are costs incurred by obligations towards the financing and
245 replacement aspects of the system. It includes interest, depreciation and
246 replacement, rents, assurances and taxes. Depreciation is a particularly
247 important aspect of fixed costs for this approach can allow for the build-up of
248 funds to replace especially larger pieces of equipment and parts in the system
249 i.e. pipes. Depreciation is way to earn back, from annual income, costs incurred
250 on the system during construction. Depreciation periods for a water system are
251 relatively long. On one hand, the technical facilities should last a relatively long
252 time (buildings and pipes); on the other hand, incomes from water sales and
253 subsidies should be realised during the entire depreciation period. While the
254 depreciation period preferably equals the expected actual technical life time of
255 the system, the depreciation periods are not necessarily equal for all the
256 components of the system. Buildings, machines, distribution network,
257 inventories all have different life times. Therefore, the costs for funding the
258 replacement reserves have to be determined separately for each component.
259 Lastly, monitoring, surveillance and training are also mainly operational
260 costs. These activities are required to continuously assess and maintain the
261 quality of the water of the source as well as during and after treatment and
262 distribution. They require skilled human resources, laboratory facilities and
263 training facilities, vehicles and sampling equipment. Some of these activities
264 might require an initial capital investment such as on-site monitoring systems.
265 A significant cost component of all these activities can be travel costs. Travel
266 might be required to and from monitoring points, remote training sessions and
267 facilities. While not conventional, costs for monitoring and assessing the
268 process of livelihood changes attributable to the intervention i.e. social
269 behavioural change, education and promotion, might also be incurred. A cost
270 component that might occur is the cost of water corruption. This activity is
271 where people illegally gain access to the distribution of the water supply such as
272 illegal and unmetered connections or inequity in distribution.

273 ESTIMATING COSTS FOR A SMALL WATER SYSTEM

274 Three costing types exist for small water supply interventions. A service
275 provider might want to 1) do a direct costing, or 2) estimate the costs as part of a
276 cost-effectiveness analysis. This chapter is about 3) estimating the costs as part
277 of a social cost-benefit analysis.
278 For guidance on direct costing, a detailed process of costing can be found in
279 the WHO publication, Costing Improved Water Supply Systems for Low-income
280 Communities, A Practical Manual (Carlevaro and Gonzalez, 2008).
281 The approach by Clasen et al. (2007) on The cost-effectiveness of water
282 quality interventions for preventing diarrhoeal disease in developing countries
283 can be followed to estimate intervention costs for a cost effectiveness analysis.
284 Estimates for cost-benefit analyses need not be as detailed as the former.
285 They can be simple unit costs as shown in Table 1 later on. The unit cost
286 approach provides flexibility when a service provider wishes to estimate
287 whether costs to install a new or upgrade an existing system are cost-beneficial.
288 Unit costs are robust cost estimations of a system that include capital costs as
289 well as recurrent costs. Data can be obtained from local sources, in particular
290 from country-specific cost summaries of previously installed schemes.
291 This section has three parts. Firstly, the costs incurred by preliminary
292 requirements to developing and installing a system. Then, costs of each activity
293 usually included in a small system are discussed. Lastly, a tabled summary
294 (Table 1) of unit costs is provided and the calculation of unit costs briefly
295 discussed. This section therefore aims at informing the reader about the full
296 costing process that eventually allows for a unit cost (e.g. cost per volume unit )
297 to be estimated.

298 Preliminary requirements

299 Costing can only begin after the physical system details are known (refer to the
300 earlier section on the implications for costing). For planned systems, an
301 inventory has to be developed to the point where specific system components
302 have been clearly identified, e.g. pipe lengths and diameter, storage tank
303 positions and sizes, etc.. For existing systems, the convenient option would be
304 to find the original technical drawings and specifications to which the system
305 had been built. This option, however, is often not available. For example, the
306 drawings might be deposited in some remote archive from where it is difficult,
307 if not impossible, to retrieve them. Even if they are retrieved, special attention
308 must be given to details of the original planning and those of the actual system,
309 to establish whether it had not already been adapted or changed since the
310 original construction.
311 It is highly recommended that costing is preceded by thorough fieldwork, in
312 close collaboration with the local community. In the absence of engineering
313 drawings, the most feasible way is to locate and map the system components
314 (increasingly convenient with GPS technology), to locate the pipe routes and
315 water connections, and also to assess the quality of the system in terms of
316 maintenance and its reliability.
317 Activity cost estimation

318 Developing the source

319 Small water systems are often supplied from groundwater or from perennial
320 protected springs. Because of its inherent characteristics, ground water in rural
321 areas is quite often considered as safe enough to be provided directly without
322 treatment i.e. with a handpump. Costs are relatively lower than with other
323 forms of supply, which makes it a popular choice with service providers.
324 Where there is no other option than to use surface water, construction of
325 impoundments in rivers and streams is mostly required (Chapter 6) to provide a
326 continuous supply of raw water throughout the year for treatment and
327 distribution. The costs of creating an impoundment in a small water supply
328 system can be a considerable proportion of the whole system cost.
329 The capital cost of groundwater sourcing are two-fold; the direct costs of
330 gaining access to an aquifer either by drilling a borehole or digging as well, and
331 the cost of lining such a borehole or well where the well has to penetrate soft
332 material in the earth. A good estimate of drill-well costs can be made, for
333 example, by using unit rates for linear metres of hole drilled and lined,
334 respectively. The unit cost here is usually capital cost per meter drilled
335 including the final finishing of the well such as casing and concrete surface
336 collar - depending on the extent of the service rendered by the drilling company.
337 The final capital cost will therefore depend on the depth of the drill-well.
338 The maintenance cost will be a percentage of the civil structure as shown in
339 the relevant section later on. Operation costs will be minimal on the well itself
340 if the well was properly installed. Operation costs around pumping will be dealt
341 with in a following section.
342 Costs to surface water sourcing will mostly be incurred by the creation of an
343 impoundment, as well as by securing the land that the impounded water might
344 cover and the land required for the sourcing activity such as a pumping station
345 and often the treatment facility. Capital costs can be estimated as cost of m3 of
346 concrete in the dam wall, per running meter of the dam wall or per m3 of water
347 stored. The latter would usually be used if the activity required purchasing of
348 land. Maintenance cost will be required for ensuring the integrity of the
349 impoundment wall, as well as for whatever sluices / valves and other
350 mechanical water outlets there might be. Maintenance cost will be a percentage
351 civil structure costs. Operation costs will be incurred by running the above and
352 will often comprise only personnel costs.

353 Storage
354 After sourcing, water usually needs to be stored, either for direct distribution or
355 pre- and post-treatment distribution. These activities require a storage tank,
356 which is usually a capital cost item. Three common storage tank types are in
357 use for small water supply systems. The smallest systems generally use
358 prefabricated glass-fibre tanks if and where these are available. These tanks are
359 available is in multiples of about 2 500ℓ or 5 000ℓ up to about 20 000ℓ. For
360 tanks from about 20 m3 or larger, tanks of reinforced concrete might be used.
361 Tanks assembled from prefabricated panels of galvanised steel are also popular
362 due their ease of construction and are available in similar volume sizes as the
363 plastic tanks.

364 Treatment
365 When water is obtained from a surface (and sometimes a ground) water source,
366 treatment is required. Depending on the quality of the source, simple
367 chlorination can be sufficient. When water is polluted with suspended solids
368 and pathogenic micro-organisms more advanced treatment is necessary,
369 including coagulation/flocculation, and filtration as well. Most of the treatment
370 items are usually capital costs to firstly install the treatment system. These costs
371 depend on degree of pollution of the source, the number and type of treatment
372 steps and the scale of the treatment. The larger the scale of the treatment is, the
373 lower the costs per m2 building area. Unit costs for different treatment steps can
374 be obtained from projects that are realised earlier in similar settings. Part of the
375 capital costs at the treatment site is the installation of a small laboratory for
376 water quality analysis, storage of chemicals, pumping stations and reservoirs.
377 Although the capital costs of treatment are normally not high compared to the
378 capital costs of transport and distribution, treatment requires considerable
379 operation and maintenance. O&M costs consist mainly of salaries for operators
380 and laboratory personnel and the costs of chemicals (such as aluminium
381 sulphate and chlorine) to be dosed during treatment. Water will be lost during
382 cleaning and backwashing of filters and disposal of the resultant sludge must be
383 organised. The loss of water (which can be up to 5-10% of the produced water)
384 represents an economic value and the sludge must be treated before disposal,
385 which represents an economic and environmental value. These costs must thus
386 be included in the O&M costs.

387 Distribution
388 Costing a distribution system is discussed in this section in the context of
389 costing mainly three diverse systems. Water can be distributed through 1) a
390 pipeline, 2) mobile units such as tanker trucks and other forms of mobile
391 vending i.e. animal drawn carts and 3) containers that people in communities
392 use to move water from the supply point and store at home.
393 Pipelines are usually capital cost items. The cost components of a pipeline
394 consist of the costs of pipes, couplings and shut-off valves. There are also the
395 earthworks needed to excavate pipe trenches, bedding for laying the pipes on,
396 backfilling the pipeline trench after laying the pipe and labour. For the smaller
397 diameters of pipes used in small systems, the capital costs are about constant
398 and mostly independent of the pipe diameter. Maintenance costs are normally
399 incurred to maintain valves. Operational costs will be incurred to fix major
400 (breaks) and minor leaks in pipelines.
401 Mobile distribution might also require considerable capital investment
402 depending on the type of system. For example, it may require investment in the
403 truck or cart and the animals. The maintenance cost will be keeping the
404 vehicles and tanks in a good mechanical state. Animals of course have to be
405 kept in a healthy state physiologically, which will incur a cost. Operationally,
406 the vehicles / animals have to be fuelled / fed.
407 A container-based distribution system requires purchasing of the containers
408 (a capital cost), and keeping the containers free from dirt and biofilm (a
409 recurrent cost item). These costs can be considerable for a poor household and
410 should be considered when attempting a cost-benefit analysis. The idea is that
411 an intervention must be optimally effective at a minimum cost..
412 Costs that is often overlooked when assessing a small system will be those
413 related to the inevitable water losses though especially the distribution part.
414 These can be seen as operational or other costs once the loss-characteristics of a
415 system are established.

416 Pumping
417 Pumping is an integral part of many small systems across the globe. Whether
418 water is pumped from the source to the treatment works or to the system, pumps
419 have certain characteristics which will enable the costing of the pumping
420 component of a small system intervention. These characteristics are best
421 determined with the help of a technician or engineer with specific knowledge in
422 this field.
423 Pump suppliers can provide an estimation of the capital as well as the
424 recurrent costs if they can be provided with information on the net power
425 delivered by the pump. This is derived from the static head, an estimate of the
426 friction head as well as the pumping rate if it is known. The pumping rate can
427 be estimated from the pipe diameter and assuming a pipe flow velocity
428 (typically between 0.6 and 1.0 m/s for small diameter pipelines). From this the
429 size of the motor to drive the pump can be derived. Such a motor can be electric
430 but would in rural areas usually be a fuel-powered motor, which has
431 implications for the recurrent costs.

432 Public source points (taps on standposts)

433 The community sources its supply from the taps at the end of standpipes, which
434 are connected to the distribution pipeline. The capital investment goes towards
435 the taps, pipework and connecting fittings, which can have a nominal size of
436 either 15mm, 20mm (the most common) or 25mm, the latter being much
437 sturdier, of course. To facilitate the filling and lifting of containers, most taps
438 are installed as part of a small concrete platform with the vertical pipe encased
439 in some form of concrete pedestal. The maintenance of the taps has proven to
440 be a substantial recurrent cost in that the taps are often not designed for the
441 many times it is opened per day and also other abuses.

442 General remarks on estimating maintenance cost

443 A longer-lifetime project often results in the replacement of more parts of the
444 installation within this period and, thus, higher maintenance costs. The planning
445 (and concurrent costing) of maintenance should identify all the activities
446 involved and the activity levels for implementation such as hours of work by
447 activity, replacement parts, repairs procedures, inputs, etc. These activities and
448 activity levels that typically would be encountered during maintenance can
449 evolve over time, making it possible to estimate an annual cost for maintenance
450 as an annual constant cost equivalent to the present value of the changing
451 maintenance costs over the use-life of the equipment. A more straightforward
452 and more generic method is to estimate the maintenance costs per year as a
453 percentage of the construction costs. The civil, mechanical and electrical parts
454 require maintenance to different extents, requiring different percentages. For the
455 specific parts the percentage has to be estimated as accurately as possible
456 together with the depreciation period. Generally, the following average
457 percentages for structures and installation costs are used to determine the annual
458 maintenance costs for the total treatment plant i.e. civil structures – X%;
459 mechanical installations – Y%; electrical installations – Z% and furnishings –
460 AA%.
461 The one important element that can make it difficult for a service provider to
462 use such straightforward approaches as described above may be the growth of
463 the population around a new system – especially in the developing world. This
464 would require constant upgrade without an incremental upgrade necessarily
465 being required. To keep an eye on effective and sustainable maintenance the
466 service provider would use a monitoring tool such as water demand.
467 The total demand to be met by a water system is a critical parameter which,
468 in a way, drives the entire cost estimate but is especially important in planning
469 and costing maintenance. In an existing system this can be directly measured
470 from a bulk flow meter (which is seldom present or working) and this value
471 should then take precedence. When this is not possible, it may be possible to
472 determine the pumping rate (by volumetric measurement of how rapidly the
473 storage tank is filled, or simply by reading the information plate on pump) and
474 determining for how many hours a day the supply pump would typically work.
475 Failing this, the water production has to be estimated from the consumer end by
476 multiplying the per capita water demand with the population. The per capita
477 demand can be estimated by counting the containers filled at a typical standpipe
478 and the population either from census data (where available) or by counting the
479 households and estimating the average occupancy per household on some
480 demographic basis.
481 To the water demand estimated in the previous paragraph must be added the
482 water lost through leaks in the pipes and at the connections. This is measurable
483 by checking the night flow, but this is less than reliable when standpipes are left
484 open during the night for irrigation or other purposes. A first estimate can be
485 made by assuming the IWA benchmark values for leakage from Loss in Water
486 Distribution Networks by Farley and Trow (2003). However small it may seem
487 at first sight, it is important to allow for some leakage – for spread-out rural
488 systems with low demand it may be a significant contribution.

489 General remarks on estimating operational costs

490 In general, operation costs are constants over the time if the prices of inputs e.g.
491 $/KWh, and activity level or output e.g. m3 of drinkable water delivered,
492 remains constant. In this case, operation costs can be estimated as a constant
493 annuity over the use-life of the equipment. If this is achievable, an annual
494 constant equivalent cost could be estimated, in the same way as for maintenance
495 costs. Parts of the operation costs are normally those required for consumables
496 such as electricity, treatment chemicals and liquid fuel for pumping stations.
497 The projected consumption of these items, as well as their prices are readily
498 estimated.
499 Much more difficult is the estimation of the cost of personnel, as they are
500 often only partially utilised in a small water supply system. Small systems may
501 only require one hour of operation per day (to stop and start a pump). Often a
502 specific person will be tasked with other community duties such as waste
503 collection. In other instances, one person will be tasked with the operation of
504 more than one small system, to which must be added the extra cost and time of
505 moving between these system. It is clear that no single algorithm could capture
506 all these permutations. There is no option but to estimate the personnel costs
507 from first principles.
508 The same arguments hold for the cost of equipment required for monitoring
509 and maintenance. A certain minimum of laboratory equipment, for example, is
510 required for monitoring, whether hundreds of only a few samples have to be
511 analysed per week. Often, the monitoring will be performed by a better
512 equipped regional laboratory to obtain some economy of scale, but at the
513 expense of transporting samples.

514 Estimating unit costs

515 Unit costs will vary between countries and will depend on the initial investment
516 (capital) costs, the recurrent costs, the life time of the system and the water
517 demand i.e. the water requirements per person per calculation period. Table 1
518 contains figures (derived from literature) for these cost components. It is
519 assumed that the various costs as discussed in the previous sections are included
520 wherever required within these figures. The unit cost for discussion in this
521 section will be expressed in US$ per cubic metre (m3) of supply water per year.
522 Table 1: Information to calculate unit costs for rural water supply systems
Capital Recurrent System Water
investment (% annual cost) lifetime demand
($ per person) (years) (Lppd)
House connection 92-144 20-40 30-50 80-120
Standpost 31-64 0-10 10-30 50-80
Handpump on drill well 17-55 0-10 10-30 20-30
Dug well 21-48 0-10 10-30 20-30
Rainwater 34-49 5-15 10-30 20-30
523
524 If house connections are used as example, then the following serves as a
525 demonstration:
526 A capital cost of $120 is assumed (from Table 1). The depreciation is linear
527 at 2.5%. The system lifetime is 40 years. Assuming an interest rate of 7.5%,
528 the fixed costs for a house connection in a small rural water system will be
529 (2.5+7.5) = 10% of $120, which is $12 per person per year. The recurrent costs
530 will be approximately 30% of $12, which is $3.6 per person per year, amounting
531 to an annual cost of $15.6 per person. Assuming a demand of 100 litres per
532 person per day (Lppd) the total annual demand is 36,500 litres which is 36.5m3.
533 The unit cost will then be $0.43 per m3.

534 BIBLIOGRAPHY
535 The chapter is based on work of a similar nature described in the following
536 literature:
537 Carlevaro, F. and Gonzalez, C. (2008). Costing Improved Water Supply Systems for
538 Low-income Communities. A Practical Manual. University of Geneva/World Health
539 Organization
540 Clasen T, Haller L, Walker D, Bartram J and Sandy Cairncross (2007). Cost-
541 effectiveness of water quality interventions for preventing diarrhoeal disease in
542 developing countries. IWA Publishing. Pp 599 – 608.
543 De Moel PJ, Verberk JQJC and Van Dijk JC (2006) Drinking Water-Principles and
544 Practices, World Scientific Publishing.
545 Farley, M., and Trow, Stuart. (2003) Loss in Water Distribution Networks. IWA
546 Publishing, London
547 Haller L, Hutton G and Bartram J. (2007). Estimating the costs and health benefits of
548 water and sanitation improvements at global level. Journal of Water and Health,
549 05.4.
550 WHO/UNICEF. (2000). Global Water Supply and Sanitation Assessment 2000 Report.
551 The World Health Organization and the United Nations Children’s Fund, Geneva,
552 Switzerland