Every reverse osmosis design starts with the water analysis. It decides the pre-treatment, the membranes, the recovery and the chemicals the plant will need. A partial analysis leads to guesswork, and guesswork in RO design usually shows up later as fouling, scaling or a plant that never reaches its output. Here is what to ask for.
Why conductivity alone is not enough
Conductivity or TDS gives a quick idea of salinity, and many people send only that. It tells the engineer roughly how much pressure the plant needs, but nothing about which salts are present. Two waters with the same TDS can behave very differently on a membrane: one may be dominated by sodium chloride, the other by calcium sulphate, which scales far more easily.
The core parameters
Ask the laboratory for a full ion analysis so that cations and anions can be balanced. A typical list includes:
- Cations: calcium, magnesium, sodium, potassium, and where relevant strontium and barium
- Anions: chloride, sulphate, bicarbonate, carbonate, nitrate, fluoride
- pH, conductivity, TDS and temperature at sampling
- Silica
- Iron and manganese, total and dissolved if possible
- Turbidity and suspended solids
- Total organic carbon or another measure of organics
From these, the engineer calculates scaling tendencies, chooses antiscalant and sets a safe recovery. Barium, strontium and silica can limit recovery even when they are present in small amounts.
Parameters that depend on the use
Some values matter only for certain applications. For irrigation, boron is critical, because many crops are sensitive to it and standard RO does not always remove enough; see boron removal. For drinking water, microbiological tests and specific contaminants named in the applicable local requirements are needed. For industrial use, the process may set limits on chloride, silica or hardness in the product water.
Iron, manganese and chlorine
Dissolved iron and manganese oxidise when they meet air or oxidants and form deposits on membranes. Wells, especially in the north of the UAE such as around Ras Al Khaimah, can carry them. If they are present, an iron and manganese removal stage goes ahead of the RO. Free chlorine, on the other hand, damages common RO membranes, so municipal water normally passes through activated carbon or dechlorination first. Ask the lab to report free and total chlorine where it is relevant.
How to take a useful sample
A perfect analysis of the wrong sample is no help. Sample from the point that will actually feed the plant, after any existing tanks if that is how the water will arrive. Let a well run before sampling so the water is representative. Use the bottles and preservation the lab specifies, especially for iron, manganese and microbiology, and deliver the samples quickly.
One sample or several?
For network water one recent sample is often enough. For wells and seawater, quality changes with pumping, season and tide. Several samples over time give a much better design basis, and the poorest result should be used for sizing.
Reading the results
Once the report arrives, check that it is complete and consistent. The sum of cations and anions should roughly balance; a large gap suggests a missing parameter or a lab error. Compare TDS with conductivity, and see whether hardness, alkalinity and pH make sense together. Values reported as below the detection limit are useful, but only if that limit is low enough for RO design.
Flag anything unusual early. A high silica result, traces of barium, or iron well above what the area normally shows all change the design, and it is far cheaper to repeat a sample than to build the wrong pre-treatment.
Record the context
Note the sampling date, the location, whether the well had been running and anything unusual. These details help the engineer interpret odd results later.
If you already have an analysis, send it with your enquiry through the quote form. If not, tell us the source and use, and AFG Dubai will suggest which parameters to request before the design starts.