We humans like order. We like putting ideas into groups just as we like putting our socks in one drawer and shirts in another. We hate being confused and nothing gives us more opportunities to be confused than the dizzying range of sensors that make measurements in water and the parameters tied to these measurements.
Take the measurement of “stuff” in the water. When we only care about how much “stuff” there is and don’t particularly care to know what that “stuff” is we are confounded by several key parameters: total solids, total dissolved solids (TDS), total suspended solids (TSS), conductivity and turbidity. How can four parameters define the one question as to the amount of “stuff” in water?
The answer is two-fold. If we stir “stuff” into a beaker of water and walk away, then two things are going to happen. Some will eventually sink to the bottom (or float) and some will stay in solution until the end of time (at least as long as the temperature and concentration remains the same). If it eventually sinks—or floats—it’s a suspended solid. If it stays in solution, then it’s a dissolved solid. This distinction doesn’t always hold. Milk is an example of a collection of suspended solids—a suspension—that does not settle out as long it doesn’t exceed its expiration date.
The distinction between the two is ambiguous and we also hate ambiguity. So the EPA, in its Methods 160.1 and 160.2, came up with the following prescription: Pass a sample through a glass fiber filter. Anything that passes through is considered dissolved and anything left behind in the filter is \suspended. Unfortunately, the methods do not specify the size of particles that pass through a glass fiber filter. Depending on the particular filter the size of particles that pass through is 0.4 to 2.0 µm (400 to 2000 nm).
This choice of filter makes sense. The reason suspended matter looks—well—suspended is that it scatters visible light. Though an oversimplification, a particle scatters the wavelength of light that is about the same size as it. The visible spectrum spans the range 430 to 700 nm. So, a solution of particles smaller than 430 nm appears clear and a solution with particle between 430 and 770 nm looks “milky,” like the glass of who-knows-what in the accompanying picture. Anything much larger than 700 nm absorbs light rather than scattering it. The question of particles between 0.7 and 2.0 µm is ambiguous. It would have been better if EPA specified an upper range of the glass fiber filter to 0.70 µm instead of 2.0 µm.
Most small particles that fall into the category of “dissolved” are ions. Ions arise when a salt dissolves into its constituent positive ions (cations) and negative ions (anions). For instance, table salt, NaCl, dissolves into Na+ cations and Cl– anions. But even this distinction is not universal. For instance, that other “table” condiment—sugar—dissolves as intact, uncharged sucrose molecules. Organic compounds whose molecules have some charge separation dissolve in water, but they are typically swamped by dissolved salts.
The only way to directly measure TDS (mg/l) is to pass a sample through a glass fiber filter to remove suspended particles, evaporate away the water and weigh the leftover solids. It’s not very practical but it’s the ONLY way to directly measure TDS. However, inasmuch, as dissolved salts make up the bulk of TDS, then measuring the conductivity of a sample of water gives us a decent approximation. Measuring this surrogate for the concentration of ions is easy. We measure the resulting conductivity of these dissolved charge carriers.
Conductivity measurement is a breeze. A conductivity probe (the contacting kind) measures the flow of current between two or more electrodes. The more dissolved ions the greater the current. The unit of measure for conductivity is Siemens/cm (S/cm). Since a Siemen is a really big unit, we prefer millionths of Siemens or µS/cm. Once we have our conductivity measurement, we only need to convert that value to the corresponding TDS value in mg/l. Wouldn’t it be convenient if all samples of water had the same correlation factor that allows us to convert conductivity in µS/cm to TDS in mg/l? We would call this factor K and the conversion would be as simple as:
Sadly, each sample of water has its own unique collection of dissolved solids so there is no one universal constant. We have to figure out what it is for every sample. Fortunately, if we know what’s in the water, we can use a value that’s “close enough” (within 10%). For instance, the value of K for NaCl in dilute solutions is about 0.45 whereas K for high concentrations of CaCl2 is approximately 0.80. (Some of this difference is due to the obvious fact that a calcium ion weighs more than a sodium one.)
In the good-enough-for-government-work world we can usually get by with a value that gets us within about 25% of the true value most of the time. That value is 0.65. So, a sample that gives us 1000 µS/cm has roughly 650 mg/l of dissolved solids. If you buy a TDS meter and it gives you readings in mg/l, rest assured it’s a conductivity sensor and it’s assuming a constant that is equal to or close to 0.65. Most analyzers (like the AM-2250 series) enable you to choose your own constant.
One more point before I close. Constants are supposed to stay constant. But this one doesn’t. As the dissolved solids concentration increases, K decreases. For very low concentrations of TDS, i.e. pure water, K increases more than linearly with increasing concentration. K becomes constant and then, at high concentrations, increases less than linear. The difference in K between one end of the concentration scale and the other can be higher than 30%. For acids the decreasing K constant actually turns negative at high concentration!
In the next white paper, we’ll tackle the other half of solids, total suspended solids, and its surrogate measurement, turbidity. It’s so exciting I’ll keep you suspended.