Why does a product that meets the TDS specification fail on site?

This answer covers products sold against a TDS and used in a customer's process: coatings, adhesives, cleaners, process fluids and lubricants. It does not cover a delivered batch that fails its own certificate of analysis or agreed specification, which is a batch nonconformance, or failures caused by equipment faults that no product change would fix.

What does a TDS-matched product failure at the plant look like?

A TDS-matched product failure at the plant looks like a product that passes every value on its technical data sheet and still does not do the job in the customer's process: a coating that loses adhesion, an emulsion that splits, a cleaner that leaves residue, a fluid that foams. The causes fall into three groups.

TDS test conditions differ from site conditions

TDS values come from specimens conditioned and tested in a controlled atmosphere. ASTM D618-21 states that the properties of plastics "are influenced by temperature and relative humidity in a manner that materially affects test results" (ASTM D618-21, 2021). The customer's substrate, water, temperature, dilution and contamination are rarely the lab's.

The property that failed was never on the TDS

A TDS lists the properties the maker chose to report. An article on supplier qualification argues that "most quality disputes trace back to an unspecified secondary parameter and not the headline assay" (Manufacturing Chemist, 2026). A matched TDS says nothing about a property it does not list.

Success was never defined

If nobody wrote down what "works" means at the plant, with a measured value, a test and a time frame, a product can meet the TDS and still be judged a failure. Typical TDS values are not a promise about use: one data sheet states that "conditions of use are beyond our control" (AC Plastics, 2026).

Which TDS test conditions differ from conditions at the customer's plant?

Five TDS test conditions most often differ from conditions at the customer's plant: the substrate, the water, the temperature and humidity, the concentration in use, and contamination. Each one changes how the product behaves without changing what the TDS says.

The substrate on the TDS is a standard test panel or specimen. The plant uses its own alloy, surface treatment and surface condition, often with oil or oxide on it.

The water used to dilute a product in the lab is usually clean. Plant water varies. For metalworking fluids, OSHA states that water with calcium and magnesium forms "insoluble soap scum when mixed with many MWFs" and that sulfates and chlorides contribute to rust (OSHA MWF manual, accessed 2026).

Temperature and humidity in the lab follow a standard atmosphere such as those specified in ISO 554:1976 or ASTM D3924-24 for paints (ISO 554:1976; ASTM D3924-24, 2024). ASTM E171-11R20 notes that end use conditions such as freezer, refrigerated or abusive storage "need to be individually established" (ASTM E171-11R20, 2020).

Concentration in use drifts. Evaporation raises the concentration of a water-mixed fluid, and OSHA links a sump 30% below the full mark to fluid loss or evaporation (OSHA MWF manual, accessed 2026). Contamination from cleaners, tramp oils, other products and process residues is absent from a lab test.

How can you tell on site whether TDS test conditions or the product caused the failure?

You can tell on site whether TDS test conditions or the product caused the failure by checking one condition at a time against the lab condition, and by testing a retained sample of the same batch. A condition gap shows up when the product works under lab conditions and fails under plant conditions; a product or batch fault shows up under both.

The table below decides whether to change a site condition, the product or the batch.

The last row decides between the product and the plant: if a retained sample of the batch passes its certificate and works under lab conditions, the cause sits in the plant conditions, not in the product.

Change only one condition per check and keep everything else at plant values, or two gaps will hide each other. Write down each reading with the date, the sample point and the method, because the same records later show whether a fix held.

How do you fix a TDS-matched product that fails at the plant?

You fix a TDS-matched product that fails at the plant by changing the condition the site check points to, retesting under plant conditions, and changing the product only if the condition cannot change. Work in this order, one change at a time, and record each result.

1. Fix the condition that is cheapest to change. A water gap can be closed by treated water: OSHA states that water for metalworking fluid mixtures "should be as pure as possible" and names deionization and reverse osmosis as ways to remove minerals (OSHA MWF manual, accessed 2026).

2. Bring concentration back into the maker's range and keep it there. OSHA recommends that actual concentration in machines be checked frequently and adjusted with pure water, concentrate or premixed fluid (OSHA MWF manual, accessed 2026).

3. Remove contamination at its source, such as a cleaner carried over from a previous stage, before judging the product again.

4. Change surface preparation or the process window, such as cure temperature or dwell time, within the limits on the TDS.

5. If the plant condition cannot change, ask the supplier for a product designed for that condition, and test it on the customer's part, in the customer's water, at the customer's temperature.

Any step that handles, mixes or heats the product follows its safety data sheet (SDS); in the United States, OSHA requires employers to keep the SDS for each hazardous chemical readily accessible to employees during each work shift (OSHA 29 CFR 1910.1200, 2024).

Which customer requirements and success criteria should be fixed before a TDS-based recommendation?

The customer requirements and success criteria to fix before a TDS-based recommendation are the plant conditions the product will meet, the properties the job depends on, and a measured definition of success. A TDS comparison means little until these are written down, because each TDS value holds for the conditions it was tested under.

Record these before you compare any TDS:

• The substrate: alloy or material grade, surface treatment, and the surface condition as it reaches the product.

• The water, if the product is diluted: source, hardness and any treatment.

• The temperature and humidity range at the point of use, by season if it varies.

• The concentration or film build in use, how it is measured and how it is kept in range.

• Known contaminants and the process stages before and after the product.

• The properties the job depends on, including secondary ones the TDS may not list, such as residue, foam, odor or compatibility with seals and other fluids. OSHA's fluid selection checklist asks whether compatibility with "machine lubricants, seals, metals, and process cleaners" was included in the selection process (OSHA MWF manual, accessed 2026).

• The success criteria: the test, the pass value, the time frame and who signs it off.

When should a plant trial replace a TDS comparison?

A plant trial should replace a TDS comparison when the site conditions differ from the TDS test conditions, when the property that matters is not on any TDS, or when a previous TDS-matched product already failed. In those cases two data sheets compared side by side cannot predict the result, because neither describes the plant.

A trial is valid when it is designed before it starts. Fix the success criteria from the list above, run the candidate on the customer's own parts under normal production conditions, and run the current product or a known reference alongside it where possible. Record the conditions during the trial, not only the outcome: water, temperature, concentration and contamination readings tell you later whether a pass or fail came from the product or from a change in the plant.

ASTM E171-11R20 says conditioning should be considered when comparing results between laboratories, "for example, supplier and customer" (ASTM E171-11R20, 2020). The same logic applies to a trial: supplier and customer must agree on the conditions before they compare results. Lubinpla reads the project's condition measurement history, product use records and work logs to fix the requirements and success criteria before it compares candidates.