By Rick Andrew
Total dissolved solids (TDS) is defined under NSF/ANSI/CAN 330 Glossary of Drinking Water Treatment Unit Terminology as, “the solids remaining when a solution is filtered through a 0.45 μm glass filter and the filtrate is evaporated and dried to constant weight at 180 °C (356 °F). TDS is expressed as mg solids per liter of filtrate as described in Standard Methods 2540.” Essentially, TDS consists of the various minerals and other constituents that are dissolved, not suspended, in water.
There are multiple treatment technologies that are capable of reducing TDS in drinking water. As such, two different NSF/ANSI Drinking Water Treatment Unit (DWTU) standards include requirements for systems making claims of TDS reduction. Systems using technology other than RO are evaluated under NSF/ANSI 42 and POU RO systems are evaluated under NSF/ANSI 58.
TDS reduction under NSF/ANSI 42
Plumbed-in and faucet mounted treatment systems, like all systems evaluated under the NSF/ANSI DWTU standards, are tested in duplicate. Initially, two test units are conditioned according to the manufacturer’s instructions. The test water contains 1,500 mg/L of TDS, which is prepared starting with tap water, using sodium chloride added to increase the TDS to 1,500 mg/L. The test units are then tested at the manufacturer’s rated service flowrate and a pressure of 60 psig. During the test, the pressure is not readjusted. The test units are operated on a 50-percent-on/50-percent-off basis for 16 hours each day, followed by an eight-hour rest period under pressure. Alternatively, a 10-percent-on/90-percent-off operating cycle may be used at the manufacturer’s discretion.
Sampling of the challenge water and the filtered water from each unit begins during the on portion of the operating cycle after 10 unit volumes of water have passed through each test unit. Afterwards, samples are collected at 20, 40, 60, 80 and 100 percent of the manufacturer’s estimated system capacity. In order to be passing, the filtered water samples must not contain more than 500 mg/L TDS. The TDS reduction requirements for treatment systems tested under NSF/ANSI 42 are summarized in Figure 1.
TDS reduction under NSF/ANSI 58
RO systems with storage tanks and automatic shut-off valves are tested for TDS reduction under NSF/ANSI 58 according to a seven-day test protocol requiring evaluation of two test systems. The test water is created starting with chlorine free deionized water. Sodium chloride is added to achieve TDS of 750 mg/L. Inlet pressure of 50 psig is set for this test and is not readjusted during the test.
The two test units are operated under a variety of conditions designed to simulate typical product usage patterns, including:
- Complete emptying of the storage tank at the sample point. This is the optimal operating scenario for systems with air pressurized storage tanks, because an empty storage tank creates the least amount of back pressure and hence, allows the highest net driving pressure and the best system performance. As the tank fills, the back pressure increases, net driving pressure decreases, permeate flow decreases and the TDS concentration of the product water increases.
- Partial emptying the storage tank. This is a less favorable operating scenario than complete emptying of the storage tank, although one that occurs in real world usage when consumers are drawing single servings of drinking water. It may require more than one partial tank draw to empty the storage tank to the point where the automatic shut-off valve turns on and allows the system to begin processing water. In any case, systems with air pressurized storage tanks will be operating under lower average net driving pressure with partial emptying of the storage tank compared to complete emptying of the storage tank.
- A two day stagnation period. This condition assesses the susceptibility of the system to membrane creep. Creep is a term that describes diffusion of TDS across the membrane from the reject side to the product side. Creep can cause significant concentrations of TDS to move into the product water under stagnant conditions, which can affect the overall TDS reduction of the system when measured according to the protocol in NSF/ANSI 58.
Differences between NSF/ANSI 42 and 58
There are obviously differences between the TDS reduction requirements for NSF/ANSI 42 and 58. The operational procedures are very different because of the unique characteristics of the technologies being evaluated. But what is perhaps more noteworthy are the differences in the test water composition and reduction requirements of the two standards.
NSF/ANSI 42 uses tap water as a basis, with sodium chloride added to make up the additional required TDS, whereas NSF/ANSI 58 requires TDS reduction test water built up from chlorine free deionized water. The reason for the more controlled water specification for RO systems is because the TDS reduction test is used to determine the recovery, efficiency and daily production rate of the RO system. Therefore, it is very important to have a highly controlled specification to allow for the greatest possible reproducibility and repeatability.
Further, NSF/ANSI 42 requires that the treated water contain no more than 500 mg/L of TDS, which is equivalent to approximately 67-percent reduction, whereas NSF/ANSI 58 requires 75-percent reduction of TDS. NSF/ANSI 58 has a higher reduction requirement because TDS reduction is used as a baseline measurement for RO system integrity and acceptability, whereas under NSF/ANSI 42 it is simply a claim of aesthetic water treatment for reduction of TDS.
And finally, the challenge level under NSF/ANSI 42 is 1,500 mg/L TDS, compared to NSF/ANSI 58, which requires 750 mg/L TDS as the challenge level. NSF/ANSI 42 is structured this way because the maximum TDS allowed in the treated water is 500 mg/L, which is the US EPA Secondary Maximum Contaminant Level for TDS. The test method paradigm is to use three times the allowable concentration in the treated water for the challenge water. On the other hand, NSF/ANSI 58 uses 750 mg/L TDS as the challenge level because it creates an established level for comparing system performance for RO systems.
More than one way to treat water, more than one way to test treatment systems
Often there are multiple technology options available to treat water. This allows professionals and end users choices to help them best fit the specific needs given the quality of the water, the volume to be treated and the treated water specifications that are required. These choices can incorporate many variables, such as initial investment, replacement frequency and costs, potential for fouling and more.
Likewise, different treatment technologies have different testing methods. These different methods can take into account variation in operational characteristics, as well as standardization of measurement to allow comparison between different system options.
About the author
Rick Andrew is NSF’s Director of Global Business Development – Water Systems. Previously, he served as General Manager of NSF’s Drinking Water Treatment Units (POU/POE), ERS (Protocols) and Biosafety Cabinetry Programs. Andrew has a Bachelor’s Degree in chemistry and an MBA from the University of Michigan. He can be reached at (800) NSF-MARK or email: Andrew@nsf.org