In a planted wastewater filter bed, sand can form part of the hydraulic and biological support; its selection needs to connect grading, bed configuration and pretreatment. The complete installation determines how wastewater enters, contacts the medium and leaves. Selecting a nominal particle size without identifying that installation is an incomplete purchasing decision.
The Netherlands has a documented research context for these systems. STOWA’s 2022 assessment considered natural post-treatment systems, including horizontal and vertical planted beds, with and without additional materials or aeration.1 These categories matter because evidence from one configuration does not automatically describe another.
This guide sets out a proposed material-selection and investigation framework. It focuses on the link between particle distribution, hydraulic operation and treatment evidence, without assigning ordinary silica sand an unverified ability to remove every dissolved contaminant.
1. Identify the bed configuration before specifying sand
Draw the water pathway first. Is water applied to the top and collected below, or does it pass through a submerged bed toward an outlet? Where are the distribution and collection layers? Which zones are intended to remain saturated?
The EPA describes subsurface systems using granular media and distinguishes their physical, chemical and biological treatment contributions.4 In a procurement document, name the actual configuration rather than use “reed bed” as a complete specification.
For the proposed selection process, annotate the drawing with media types and layer functions. Sand in the treatment layer and gravel around a collection pipe are not interchangeable quantities. A correct sand delivery cannot repair an unspecified drainage arrangement.
Check whether the project treats raw wastewater, pretreated wastewater or polishing effluent. That position in the treatment train belongs beside the material requirement. A medium selected for one position cannot be assumed suitable for another without examining the incoming load.
2. Separate a hydraulic function from a removal mechanism
The medium contributes a water pathway and surfaces within the system. However, a treatment result can arise from more than particle capture alone. Microbial processes, chemical interactions and system conditions may all matter.
The EPA design manual discusses the different constructed-wetland treatment processes and their relationship with configuration.2 Keep this distinction in the material enquiry: filtration of suspended material and removal of a dissolved constituent are different claims requiring different evidence.
For a proposed assessment, make a table of target parameters and assigned mechanisms. If the target is a dissolved compound, do not write “retained by the sand” without identifying the demonstrated process. If an added sorbent is responsible, identify it separately.
This also helps interpret product descriptions. A report about a specially amended medium does not qualify plain silica sand for the same task. The specification should distinguish the ordinary granular skeleton from any additional treatment material.
3. Ask for the complete grading and preparation condition
A name such as “washed sand” does not provide the full particle distribution. Obtain the required curve and the limits relevant to the selected design. Ask whether the specification applies to delivered material or to material after a prescribed preparation.
In the proposed comparison, assess the candidate and reference with a common method. Retain the fine fraction in the reporting rather than show only the prominent commercial size band. Differences hidden outside that band may matter to the material evaluation.
If the project specifies effective size or another distribution indicator, retain its definition. Do not substitute a convenient average without explaining the change. The full curve provides context that a single number cannot.
Keep grading and cleanliness as separate checks. Washing can remove part of an unwanted fraction, but it does not prove the material has become the required hydraulic medium. Test the prepared state actually intended for installation.
4. Measure hydraulic behaviour instead of importing a table value
Published values can support preliminary comparisons, but project design needs material-specific evidence. The EPA’s technology assessment explains that hydraulic conductivity and porosity should be determined for the selected medium before design, rather than treating indicative tables as a final certificate.2
For the proposed laboratory comparison, state specimen preparation, water conditions, geometry and the measurement procedure. A tightly packed sample and a loosely placed sample should not be compared without identifying that difference.
Report the measured value and its units clearly. Hydraulic conductivity, permeability and a measured discharge are not interchangeable labels. The laboratory should identify which quantity the procedure actually determines.
Repeat the measurement when preparation variation could affect the selection. A single favourable result cannot reveal the spread between prepared portions. Use the distribution of results to decide whether the candidate is sufficiently consistent for the adopted design calculation.
5. Distinguish surface loading from pore-water movement
A simple fictional example clarifies a common unit confusion. Applying 10 m³ per day to 100 m² gives an average surface hydraulic loading of 0.10 m per day. That number describes volume per plan area and time; it is not automatically the velocity of water through every pore.
Suppose an idealised saturated bed has 100 m² area, 0.6 m depth and 0.35 water-filled porosity. Its nominal pore-water volume is 21 m³. Dividing by 10 m³ per day gives 2.1 days of nominal residence time.
These calculations are bookkeeping for an explicitly simplified example, not a wetland design recommendation. They assume a particular saturated volume and do not establish that all water follows the same path. An intermittently drained vertical bed cannot be described uncritically by that saturated-volume assumption.
Use such calculations to check units and expose assumptions. The result is useful precisely when the project team can say which area, depth, saturation condition and water volume were used.
6. Investigate preferential pathways rather than average them away
An installation can have the intended total inflow while distributing it unevenly. A short pathway and a poorly contacted zone may coexist. A calculated average residence time does not reveal that pattern.
In the proposed investigation, inspect distribution, collection and water levels before blaming the sand. Compare observations at multiple relevant locations, using a monitoring plan appropriate to the system. Map persistent surface accumulation rather than record only a yes-or-no ponding result.
Where a specialist proposes a tracer test, define its question and approved method. Interpret the measured passage distribution rather than use only the mean. An early passage can matter even when later water raises the average.
Keep the investigation connected to the drawing. A material change that increases conductivity everywhere may not correct an inlet layout that sends water preferentially along one side. Diagnose the pathway before choosing the intervention.
7. Put pretreatment alongside the clogging assessment
The EPA manual discusses solids accumulation and clogging in subsurface beds.2 This makes the incoming condition part of the media assessment, rather than treating clogging as a property of the sand alone.
For the proposed audit, document the treatment upstream and its actual performance. A bed receiving a new solids load after a pretreatment fault is not operating under the same condition as its qualification trial.
Compare incoming suspended-solids measurements with hydraulic observations over the same period. Keep exceptional discharges identifiable. If a load increase coincides with a change in drainage, investigate that relationship without assuming it proves the only cause.
A purchasing response should follow the diagnosis. Replacing the medium without restoring the intended upstream condition can repeat the problem. Conversely, a properly operating pretreatment system does not eliminate the need for an appropriate bed distribution and installation.
8. Translate concentration into incoming mass
In a fictional example, 10 m³ per day containing 50 mg/L of suspended solids carries 0.50 kg per day. The conversion uses 10,000 litres per day and gives 500,000 mg, or 0.50 kg.
At the same concentration, doubling the flow doubles incoming mass. At the same flow, doubling concentration also doubles mass. Tracking only concentration can therefore conceal a significant loading change.
Do not assume all incoming mass is retained in the bed. Compare measured inlet and outlet streams and account for any other relevant pathway before calculating accumulation. Sampling timing should be appropriate to a variable flow.
Use the calculation as a proposed operational comparison, not as a clogging prediction. The time at which hydraulic behaviour changes depends on more than incoming mass alone. Its value is to distinguish material quality from a changed duty.
9. Avoid a universal promise for dissolved phosphorus
STOWA’s published 2005 investigation specifically examined phosphorus removal in planted filters and identified predictability and achievable effluent concentrations as questions requiring evaluation.3 This is a reason to request demonstrated performance for the chosen system rather than assign a universal removal rate to sand.
For the proposed qualification, distinguish particulate phosphorus from dissolved forms in the measurement plan. A reduction associated with captured particles does not prove equivalent removal of dissolved phosphorus.
If an added reactive material is proposed, ask how its role, capacity and replacement requirement are demonstrated. Keep its specification separate from the silica sand fraction. A mixture containing it is a different material condition.
Do not translate an initial successful measurement into an indefinite lifetime. Compare performance over the relevant operation period and water conditions. The procurement requirement should identify what evidence is needed before a long-term treatment claim is accepted.
10. Examine micropollutant evidence at the system level
The Dutch STOWA assessment compared natural treatment approaches for organic micropollutants and called for further measurement and pilot work where year-round performance remained unconfirmed.1 It does not establish that an ordinary bag of silica sand removes all pharmaceutical residues.
For the proposed evidence review, list the measured compounds rather than report only “micropollutants removed”. Identify sampling conditions, the full installation and any additional material or aeration. An average across compounds can conceal very different individual outcomes.
Separate a measured reduction from an explanation of its mechanism. A lower outlet concentration does not by itself identify whether the process was degradation, transfer or another pathway. The study’s evidence should support the interpretation used.
For a material purchase, ask which characteristics of the sand are genuinely specified by the selected system. Do not use system-level treatment results as a sand-only performance certificate. The boundary protects both the technical decision and the clarity of the supply request.
11. Compare performance under the relevant operating conditions
A short trial represents the conditions it actually experienced. If the planned installation faces variation in water composition, flow or season, define which of those conditions the qualification needs to examine.
For the proposed pilot, select comparable inlet and outlet sampling over each relevant condition. Document operating events beside the measurements. Avoid comparing an inlet peak with an unrelated later outlet sample and calling the difference removal.
Review hydraulics and water-quality performance together. An acceptable laboratory grading does not establish acceptable field distribution, while visually satisfactory drainage does not establish the required effluent quality.
Define the decision rule before the results arrive. If performance varies, identify whether an upstream change, operating adjustment or alternative treatment stage is needed. Changing sand should be an evidence-based response to a material-related finding.
12. Turn the accepted trial into an installation specification
A useful specification connects the selected configuration, layer position, required grading, preparation and measured hydraulic indicators. Include how the delivered material will be checked and the installation condition in which it is intended to operate.
Separate the delivered-medium certificate from field commissioning. The first describes material measurements; the second checks the assembled system. Neither should silently stand in for the other.
For maintenance planning, identify access and the observations that trigger investigation. Keep the response appropriate to the actual configuration rather than import a pressure-filter cleaning routine into a planted bed.
The silica sand page helps describe a material enquiry. For Dutch planted-filter projects, the strongest qualification connects sand distribution, measured hydraulics, pretreatment and system-level evidence. That creates a clear purchasing decision while preserving the distinct roles of the medium, biology and any additional treatment stage.
13. Check a hydraulic calculation’s geometry
For a fictional horizontal-flow comparison, a rectangular bed 10 m wide and 0.6 m deep has a nominal transverse area of 6 m² when that depth is saturated. Its plan area is a different quantity. Substituting the plan area for the transverse flow area in a horizontal hydraulic calculation changes what the calculation represents.
Likewise, specify the length over which a water-level difference is measured. A 0.10 m difference over 20 m gives a dimensionless gradient of 0.005. It does not mean a slope of 0.10 or a head difference of 5 mm.
Ask the designer to identify these quantities on the drawing used for the adopted calculation. This unit-and-geometry check does not establish treatment capacity, but it can expose an input error before the material requirement is issued.
Keep time units consistent too: a conductivity expressed in metres per day cannot be inserted directly into an hourly-flow calculation. Convert the quantities before comparing the alternatives.