How to Remove Salt from Borehole and Well Water: A Treatment Decision Guide

Possible causes include a naturally saline aquifer, seawater intrusion, saline water drawn in as pumping changes groundwater levels, or a particular dissolved ion problem. A salty taste alone does not identify which mechanism is present. Compare laboratory results with the well location, pumping history and any previous analyses. The US EPA's well-testing guidance lists sodium, chloride and total dissolved solids (TDS) among tests relevant to salty taste or possible seawater influence.

A single sample is a starting point. Ask for historical or seasonal analyses when the source changes with pumping, rainfall or the dry season. Confirm where the sample was taken: untreated well water, a storage tank and an existing treatment outlet are different design inputs. The well-water testing checklist gives buyers a fuller sampling sequence.

Specify the final use before setting a product-water target. Drinking water, food preparation, irrigation, boiler makeup and industrial process water have different quality and hygiene requirements. A reported TDS target alone may omit microbiology, hardness, silica, specific ions or the condition of the storage and distribution system.

For drinking-water projects, use the applicable destination-country requirements and a complete water-safety approach. RO removes a dissolved-salt load, but potable supply may also need conditioning, disinfection, protected storage and verification at the point of use. The WHO drinking-water guidelines provide a risk-management framework; they do not replace local rules. For boiler duty, use the boiler manufacturer's feed-water specification and the boiler feed-water guide. For highly purified uses, the single-pass versus double-pass RO guide explains why a second pass may be considered.

Send the laboratory report, sampling date, well location and design conditions together. A useful engineering request includes at least the following groups; the exact panel depends on the source and end use.

Do not assume conductivity and TDS are interchangeable without stating the conversion basis. Iron and manganese need special attention: the DuPont FilmTec technical manual explains how oxidation can turn dissolved metals into particles that foul RO membranes. The correct treatment route depends on measured concentrations and operating conditions. See the iron and manganese well-water guide for that separate problem.

Media filtration removes particles; a conventional softener exchanges hardness ions. Neither is a substitute for removing bulk sodium and chloride from a saline well. Nanofiltration can be useful when hardness or divalent ions dominate and only partial desalination is required. BWRO is the usual membrane option when the project needs broader dissolved-salt removal. An older WHO treatment-methods annex, from the fourth edition with the first and second addenda describes the difference between RO and NF; the final process still depends on the water report and product target.

The well-water treatment and softening guide shows where these units sit in a treatment train. If the source is true seawater rather than brackish groundwater, the selection changes; the SWRO versus BWRO comparison explains that distinction.

Start with the treated-water demand at the use point, not the nameplate flow of a pump. Then specify how many hours per day the RO can operate. A simple preliminary calculation is:

Required average RO product flow (m³/h) = treated-water demand (m³/day) ÷ operating hours (h/day)

For a site needing 12 m³/day of treated water, the minimum average product flow changes as follows. These are planning calculations, not a KONCHE equipment rating or a performance promise.

The water balance used here is feed = product ÷ recovery and concentrate = feed − product. The 65% figure is an arithmetic illustration, not a default recovery for a well. Values are rounded. The sustained well yield, pretreatment backwash and any other site demand must be checked separately; do not infer that the well can deliver the calculated feed flow. For the RO operating principle, see the industrial RO guide.

The hourly requirement also changes when the site has short peaks, planned downtime or a duty/standby requirement. Calculate a demand profile rather than adding an arbitrary percentage to the table. Ask whether the buyer means 1,000 LPH of RO product water or 1,000 LPH of feed water; these are different specifications.

Storage bridges the gap between RO production and use. Its usable volume depends on the hourly demand profile, hours when RO is off, required reserve, storage hygiene, and whether a standby train is available. For example, a site that consumes 12 m³ evenly across 24 hours uses an average 0.5 m³/h. If a 1.5 m³/h RO runs for only 8 hours, the product tank must cover the other 16 hours of demand, plus any project-required reserve. The first-pass off-hours volume is 8 m³ (0.5 × 16); real design must also account for tank minimum level, peak demand, cleaning or outage periods and water-quality control. This is not an automatic recommendation for an 8 m³ nominal tank.

Separate feed storage, treated-water storage and distribution pressure in the quotation. For potable use, a tank and downstream pipework can introduce microbial risk even when RO permeate meets its chemical target. Specify cleaning, protection and disinfection at the appropriate points.

The source report decides the train. Avoid a fixed drawing presented as suitable for every borehole.

The 1,000 LPH Egypt brackish-well RO project design provides a concrete example: its proposal specified sand filtration, activated carbon, automatic softening, 5 µm cartridge filtration, single-pass BWRO, storage and final UV treatment. The design basis included conductivity up to 5,000 µS/cm and hardness up to 800 mg/L as CaCO₃, with a proposed recovery of 60–65%. These are proposal specifications for that project, not measured operating results or a universal KONCHE configuration. A different well may need a different train.

Higher recovery reduces the volume of concentrate for a given product flow, but it also raises salt concentration on the membrane concentrate side. That can increase scaling or fouling risk and may require a different membrane arrangement or pretreatment. The practical recovery is set through water chemistry, membrane projection, source-water variation and an acceptable cleaning regime. The DuPont FilmTec manual treats recovery and pretreatment as related design decisions.

Ask for the proposed product, feed and concentrate flows at the stated design temperature and water analysis. Also identify where concentrate will go. Local discharge permission, sewer acceptance, evaporation or another route depends on the destination and must be resolved for the actual site; this article does not prescribe one route globally.

An EPC comparison is easier when each supplier responds to the same design basis:

1. A dated laboratory analysis, source variation and sustained well yield.
2. Product-water specification for every use point, plus daily and peak demand.
3. Operating hours, outage reserve, storage and duty/standby expectations.
4. Pretreatment stages, the design condition that triggers each stage and monitoring points.
5. Proposed product, feed and concentrate flow; recovery and membrane-projection assumptions.
6. Temperature and power basis, pump and membrane arrangement, controls and instrument scope.
7. Concentrate route, backwash and cleaning wastes, installation interfaces and exclusions.
8. Acceptance tests, consumables, spare parts, documentation and commissioning scope.

The single- and double-stage RO product range shows the equipment family; a project quotation should still be tailored to the well and final use. To review similar published projects, visit the KONCHE case library.