Two percentages answer different questions
Recovery asks what fraction of feed water becomes product water within a defined measurement boundary. TDS rejection asks how much a dissolved-solids concentration reading falls across the stage being evaluated. Both may be percentages, but their numerators, denominators and meanings differ. A high recovery number is not a water-purity result, and a high rejection number does not tell you how much reject water was produced.
This educational guide separates the arithmetic so that the metrics can be recorded without confusion. It does not set a required operating recovery or rejection threshold for a particular membrane. Those limits and the appropriate testing procedure belong to the manufacturer. The RO/DI production calculator supports collection-based volume recovery, not a feed-versus-permeate TDS-rejection calculation or a purity certification.
Define the water-collection boundary
For a simple steady collection with no other streams or inventory changes, feed volume equals product plus reject volume. Suppose a simultaneous interval produces 2 L of product and 6 L of reject. The represented feed is 8 L, so recovery is 2 divided by 8 times 100, or 25 percent. The reject-to-product ratio is 6 divided by 2, or 3:1.
The denominator for recovery is 8 L, not 6 L. Dividing product by reject would give about 33.33 percent, but that is not recovery as defined here. Keep those labels explicit. If flushing, storage or another flow lies outside the collected interval, the simple balance may not describe the full session. A steady sample and total water consumption are different boundaries, not rival calculations of the same record.
Define the concentration comparison separately
A common TDS-rejection expression is (feed reading − permeate reading) divided by feed reading times 100, using comparable positive feed and nonnegative product readings on the same measurement convention. For an invented feed reading of 300 ppm and RO-permeate reading of 9 ppm, the calculation is (300 − 9) ÷ 300 × 100 = 97 percent. Those readings are independent of the example collection volumes.
The arithmetic requires a nonzero feed denominator and a meaningful comparison. A feed reading of zero cannot produce a valid percentage through division by zero. An RO-permeate reading above the feed reading would produce a negative calculated result, prompting review of the measurements and conditions rather than an automatic diagnosis. Do not hide such a result by taking an absolute value or replacing it with zero.
Keep RO permeate distinct from final DI output
If the question is membrane rejection, identify the appropriate RO-stage sample rather than automatically using a final DI reading. A downstream stage changes what the sample represents. A near-zero final reading does not by itself quantify the preceding membrane's performance. Label sample locations in the record so a water-quality comparison remains attached to the actual stage being evaluated.
Follow the instrument and system instructions for obtaining comparable readings, including any required operating state and temperature treatment. The term TDS itself can involve an instrument's conversion convention rather than direct weighing of every dissolved constituent. Do not mix two unrelated display conventions into one precision percentage without checking their comparability. A calculator formula cannot repair a sample taken from the wrong point or a mislabeled instrument reading.
Demonstrate why the percentages are independent
Retain the invented 2 L product and 6 L reject collection: recovery remains 25 percent. If comparable concentration readings were instead 300 ppm feed and 30 ppm RO permeate, rejection would be 90 percent. Changing the concentration inputs changes the rejection calculation without changing the collected volume balance. A favorable-looking water-use ratio therefore cannot establish the concentration result.
Conversely, an invented collection of 2 L product and 2 L reject has 50 percent recovery, while the same 300 and 9 ppm readings still give 97 percent rejection. These paired examples are deliberately separate scenarios, not instructions for changing restrictors to obtain either result. Actual system behavior links conditions in ways this arithmetic does not model, so do not treat numerical independence as permission to adjust the hardware freely.
| Metric | Inputs | Result |
|---|---|---|
| Water recovery | 2 L product / (2 L product + 6 L reject) | 25% |
| Reject-to-product ratio | 6 L reject / 2 L product | 3:1 |
| TDS rejection | (300 ppm feed − 9 ppm RO permeate) / 300 ppm feed | 97% |
Volumes belong to one simultaneous collection; concentration readings belong to explicitly identified sampling stages. These examples are not operating targets.
Project water quantities without making a quality claim
At an observed reject-to-product ratio of 3:1, a 20 L product target projects 60 L of reject and 80 L represented feed under the same steady-rate assumptions. The recovery remains 20 divided by 80, or 25 percent. Enter the simultaneous collection and desired product volume in the RO/DI tool to reproduce these quantity outputs. No TDS data are involved in that computation.
The projection does not include every startup or maintenance loss unless the chosen measurement boundary does. Nor does it assess whether reject water is suitable for another use. Keep any reuse decision separate from the volume estimate and applicable guidance. The output's label should describe projected water quantities, not efficiency in removing every contaminant or approval to use the product for aquarium preparation.
Avoid optimizing one number in isolation
Do not alter pressure, temperature or a flow restrictor simply to increase the recovery percentage shown by a worksheet. The manufacturer may impose limits and maintenance requirements that a volume ratio cannot evaluate. Likewise, a high rejection estimate from one pair of readings is not a complete treatment-performance test. System condition, sampling and intended use require their own appropriate assessment.
A useful log puts collection metrics and concentration metrics in separate groups. Include model, operating state, times, sample locations and measurement methods. If a result changes, compare like with like before assigning a cause. An unexplained decline is a reason to seek model-specific guidance, not an opportunity to invent a universal acceptable threshold or apply a correction from an unrelated commercial system.
Use the production tool for the metric it actually supports
The production calculator calculates recovery only when reject volume is supplied for the same collection interval. A blank optional reject field leaves recovery unavailable; it does not mean zero reject water or 100 percent recovery. A recorded zero reject volume is a different input, which needs particular care in interpretation and is not an endorsement of any system configuration.
There is no TDS-rejection mode in this tool. Use the definitions in this article to understand a properly obtained concentration comparison, while following the relevant manufacturer and instrument procedures. The final distinction is simple but important: recovery is a water-volume fraction, rejection is a concentration reduction, and neither should be promoted into an unsupported guarantee about all aspects of water quality or system operation.