A household biosand filter and an industrial sand filter both contain granular material, but that shared component does not make them interchangeable. Their operating pattern, treatment objective, maintenance and performance evidence can differ substantially. Choosing sand begins with identifying the actual filter design, not with assuming that any product labelled “filter sand” fits every water application.
A trial around Tamale, Ghana, assessed a plastic household biosand filter and its drinking-water and health outcomes.1 Its evidence concerns that particular technology.
This guide proposes a practical framework for distinguishing systems and preparing a material qualification request. The aim is to connect sand characteristics with the selected design while keeping treatment evidence attached to the technology that produced it.
1. Start with the water use and the complete system
Write the intended use before choosing a size range: household drinking water, process water, irrigation or another application. These uses create different quality requirements. “Clear water” is an observation, not a complete treatment specification.
For drinking water, evaluate relevant microbiological hazards and the complete treatment and handling chain. WHO’s household-treatment framework distinguishes performance against different pathogen groups rather than treating one percentage reduction as a universal safety statement.3
For an industrial process, identify the parameter that threatens operation: suspended solids, a particular size fraction or another measured requirement. Some processes require treatment beyond ordinary granular filtration. Define that requirement before assigning the sand a function.
In the proposed selection worksheet, place three items together: incoming water characteristics, the quality required at the point of use and the barriers responsible for closing the gap. This makes it harder to ask a sand bed to achieve an unassigned or unverified treatment task.
2. Understand what “biosand” identifies
A household biosand filter adapts slow sand filtration to intermittent domestic use. CAWST’s manual describes a device with filtration sand, supporting gravel, an outlet and a protected water inlet, with biological activity contributing to treatment.2 The term describes a functioning system, not a special chemical grade of silica.
A bag of sand does not contain the completed performance of that system. Geometry, preparation, installation and operation all belong to the evaluated design. Replacing its sand with a convenient industrial grading changes an important component and requires assessment.
Conversely, a pressure filter containing silica sand does not become a household biosand filter simply because microorganisms can be present in water. Identify the equipment and intended treatment mechanism explicitly.
For procurement, request the design reference and its material specification before requesting a price. If the proposer cannot name the system or explain which sand requirement applies, the purchase description is not yet sufficiently defined.
3. Distinguish operating patterns before comparing capacity
Intermittent household use and plant operation demand different capacity descriptions. An instantaneous outlet flow does not, by itself, establish the volume that can be treated acceptably each day. Filling, pauses, user availability and the accepted operating schedule all matter.
For an industrial filter, flow capacity also belongs to a specified operating point and maintenance regime. A nominal pump flow does not establish that the bed receives a uniform flow or that its treatment objective is met throughout a run.
Build a proposed demand timeline instead of comparing headline litres per minute. Show when water is needed, when the device operates and how treated water is stored. Identify peak demand separately from daily volume.
Do not increase household capacity by simply adding water faster or enlarging every dimension proportionally. An institutional intermittently operated slow sand system has its own technical manual; CAWST and Samaritan’s Purse publish a separate iSSF design resource.5 Larger use deserves a designed system, not an improvised extrapolation.
4. Specify grading as a distribution, not one grain size
A description such as “fine silica sand” leaves the distribution unresolved. A nominal size can conceal different proportions of coarse particles and fines. Ask for the full required grading and the preparation condition in which it applies.
In the proposed qualification, compare a representative sand sample with the selected design’s specification. Keep the incoming product and the prepared filtration material distinct. Sieving and washing may change the distribution, so a certificate for the incoming bulk material does not automatically describe the installed bed.
For a simple fictional illustration, two products can both include a 0.4 mm fraction while one has substantial material below it and the other contains mainly larger grains. The shared size says little about how their complete distributions compare.
If the design uses effective size or uniformity coefficient, retain those defined indicators alongside the curve. Do not replace them with an arithmetic average selected for convenience. Ask the laboratory to use the agreed method consistently for the reference and candidate materials.
5. Calculate preparation yield before estimating material purchases
Household filter programmes may prepare sand before installation. Purchasing enough material therefore requires more than multiplying the number of filters by their installed sand mass. Account for what the selected preparation rejects.
Consider a fictional plan for 100 devices requiring 20 kg of prepared material each. The installed requirement is 2,000 kg. If a trial preparation yields 80% usable material by dry mass, the corresponding incoming requirement is 2,500 kg before any separately planned handling reserve.
At a measured yield of 60%, it becomes about 3,333 kg. Those values are arithmetic examples, not a sand requirement for a CAWST design. They show why a cheap incoming tonne can have a different cost per usable tonne.
Measure yield on representative material using the intended preparation method. Separate deliberately removed coarse particles from fines and handling losses. The result helps compare supply options without assuming all delivered material will enter the filters.
6. Preserve the distinction between clean-looking and treated water
Appearance can help users notice changes, but it cannot establish microbiological performance. A clear sample and a cloudy sample do not, by themselves, reveal which relevant pathogens remain.
The Tamale trial measured microbiological and water-quality outcomes rather than relying solely on appearance. It also observed recontamination during storage after filtration.1 This is a direct reason to evaluate the outlet and the water actually consumed separately.
For the proposed implementation check, define where each sample is taken: source, device outlet, storage vessel or point of consumption. If the outlet improves but stored water deteriorates, changing sand may fail to address the actual weakness.
Maintain clear labels for different measurements. Turbidity and a microbiological indicator provide different information. A favourable result in one should not silently substitute for the other in the acceptance decision.
7. Read removal percentages with their starting values
A reduction percentage is relative to the incoming concentration. In a fictional calculation, reducing 1,000 units to 30 gives 97% removal. Reducing 100 units to 3 also gives 97%. Equal percentages therefore do not imply equal remaining concentrations.
For a study comparison, keep the result attached to the evaluated device, incoming water, operating conditions and sampling arrangement. A percentage quoted without those details cannot qualify an unrelated sand delivery.
When the laboratory reports logarithmic reduction, compare results using its stated calculation and treatment of values below detection. Zero organisms observed in a particular tested volume is not the same statement as absolute absence in all water.
For qualification, ask both what reduction was measured and whether the resulting quality satisfies the selected requirement. The two questions are complementary. A large relative improvement can coexist with a remaining treatment need.
8. Keep biological operation attached to the design
CAWST’s manual includes biological-layer development, operation pauses and maintenance as parts of biosand use.2 These features should remain connected to the applicable design instructions, rather than become generic settings for all granular filters.
In the proposed comparison, note whether the system is newly installed, operating under established conditions or recently maintained. Comparing these states without identifying them can make a material difference appear larger or smaller than it is.
If water use changes, discuss the effect with the programme’s technical operator before changing the schedule. More frequent filling is an operating change, not merely a convenient way to obtain more daily water.
For a material trial, keep the operational state comparable between the reference and candidate units. Otherwise, the trial may compare different operating histories as much as different sands. This is particularly important when interpreting an early result.
9. Do not transfer industrial backwash instructions automatically
The maintenance method belongs to the system. A pressure filter’s flow reversal or air-assisted cleaning sequence cannot be assigned automatically to a household biosand device. Identify the permitted intervention from the actual design instructions.
For a household programme, users need a procedure they can recognise and carry out correctly. Demonstrate it on the selected device and verify their understanding. A technical document that is appropriate for plant staff may be unsuitable as the household instruction.
In the proposed follow-up visit, investigate a slow outlet by checking the device’s operational condition and approved maintenance history. Avoid immediately replacing the medium without understanding whether the reported condition is expected or abnormal.
Keep replaced material and disturbed components identifiable when technical investigation is needed. A change made without noting the previous state can eliminate the evidence required to understand the problem. Maintenance should restore the selected design, not create a different filter without evaluation.
10. Include storage and dispensing in the assessment
Household treatment does not end at the outlet. Collection, storage and dispensing determine what happens between the device and drinking. WHO’s household-treatment approach explicitly includes safe storage and contamination during collection, transport and use.4
In the proposed visit, observe the actual container and dispensing practice rather than ask only whether the family owns a lid. Identify whether hands, utensils or untreated water can contact treated water. Keep the observation specific to the household situation.
Where sampling is planned, take paired outlet and stored-water measurements under a defined schedule. The comparison can distinguish treatment performance from deterioration after treatment. It should not be interpreted as proof of a cause without inspecting the handling pathway.
This creates a useful purchasing boundary: a sand supply specification can address the medium, while programme instructions address storage and use. Asking sand to compensate for an unsuitable storage practice confuses two separate intervention points.
11. Compare programme costs with the whole operating requirement
A device purchase price is only one cost component. For a proposed programme, include material preparation, construction or equipment, transport, installation, user instruction, follow-up and the measurements needed to assess performance.
Use a consistent unit for comparing options, such as cost for a defined delivered service over an agreed period. Do not compare the price of an uninstalled household device with the installed cost of a plant treating a different water volume and quality objective.
Separate measured costs from assumed ones. If usable sand yield has been tested but follow-up frequency is only planned, label each accordingly. A transparent calculation can then be updated when the pilot provides actual observations.
The decision should consider whether the operating requirements can be maintained locally. A technology that works in a closely supervised test requires an implementation plan for ordinary use. The relevant question is whether the complete service remains workable after initial installation.
12. Build a pilot that can answer the procurement question
A useful proposed pilot compares the selected design and operating procedure with defined water and material conditions. Establish its questions before installation: material preparation yield, compliance with grading, device operation, outlet quality and quality after household handling.
Record deviations that alter interpretation. A changed water source, different preparation method or major maintenance event can change what the measurements represent. Keep these events beside the affected results rather than hide them in a separate narrative.
Avoid making a health-effect promise from a small material trial. An implementation pilot can answer practical qualification questions without being a clinical study. Use published research to understand technology evidence and use the pilot to test the defined local implementation.
For Ghanaian projects, the silica sand page helps frame a material enquiry. The stronger request names the system, its grading and preparation requirements, the intended water use and the evidence needed for acceptance. That is how household biosand and industrial filtration can be distinguished without confusing a shared raw material with a shared treatment performance.