Wink of Knowledge: Mineral content of water

The mineral content of water plays a key role — both in terms of beverage quality and industrial processes. While it often causes problems such as calcification or corrosion in industry, it promotes health and influences taste in beverages. Whether it’s a bug or a feature, we show how mineral content can be reliably determined using density measurement.

Why this test?

Interest in analyzing drinking water probably goes back to the earliest days of human history. While the central question used to be, “Is the water even drinkable?”, today the focus is increasingly on optimization: “How can I improve water specifically for my individual needs?”

A key and frequently studied property is the mineral content of water, which can be used to determine water hardness, among other things. Minerals such as fluoride, sodium, calcium, and magnesium are essential for numerous bodily functions—for example, for teeth, bones, nerves, and muscles. At the same time, these minerals can also have undesirable effects in everyday life, such as calcification or corrosion in household appliances such as coffee machines or washing machines.

In the beverage industry, the taste-influencing effect of certain ions (e.g., chloride or sodium) also plays a role. In beer brewing, the chemical properties of minerals during the mashing process are also crucial—further information on this can be found at brunwater.com, for example.

Numerous methods are available today for determining mineral content: from simple test strips and electrical measurements of free ions to high-precision laboratory analyses such as mass spectrometry. One method that is less well known but very promising is density measurement, which will be examined in more detail in this article.
Density measurement is already well established in the brewing process and is therefore particularly suitable. It not only allows seasonal fluctuations in the mineralization of the water to be monitored, but also directly provides the starting value for the subsequent extraction and fermentation processes – both of which are also monitored by means of density measurement.

Results

Various commercially available mineral waters and tap water from different regions were examined. The measurements were carried out using both our DLO-M2 and the high-precision laboratory device DSA 5000 M from Anton Paar. The reference values included information provided by the manufacturers and publicly available sources such as trinkwasser.ch for Swiss tap water.

The existing concentration model of the DLO-M2 for “Total Dissolved Solids” (TDS) was used to convert the measured density values into mineral content.

FIgure 1: Mineral content of different types of water. Orange bars represent laboratory analysis, gray and blue bars represent mineral content based on density measurements with the DLO-M2 and DSA 5000 M, respectively.

The results show a very good correlation between the measurements taken by the DLO-M2 and those taken by the laboratory device (gray and blue bars), which underlines the high measurement accuracy of the DLO-M2. However, there are certain deviations compared to the actual mineral content. These can be explained in part by natural fluctuations in tap water. In addition, the physical measurement value—whether density or electrical conductivity—also depends on the chemical composition of the dissolved minerals. Factors such as charge number and molecular mass influence the measurement. Without precise knowledge of the composition, an assumption must therefore be made about a “typical” mineral distribution.

One way to estimate the relationship between the physical measurement and the actual mineral content is to perform a correlation analysis between the measurements and the manufacturer’s specifications:

Figure 2: Correlation between density measurement in mg/ml and mineral content according to laboratory analysis, also in mg/l

For the samples examined here, the conversion factor between density [mg/ml] and mineral content [mg/ml] is 0.936. Conversely, this means that an increase in density of 1 mg/l compared to distilled water can be expected to result in a mineral content of approximately 1.07 mg/l. The graph also shows that the correlation works better at low mineral content than at high content.

Figure 3: Correlation between conductivity measurement in µS/m and mineral content according to laboratory analysis in mg/l

Overall, this results in a very high correlation value of R² = 0.9889 (Figure 2). A comparable analysis using laboratory measurements of electrical conductivity yields an R² of 0.9904 – almost identical to the density measurement (Figure 3). From a measurement technology perspective, density measurement is therefore a reliable alternative to classic conductivity measurement.

Finally, seawater from the Mediterranean Sea was also examined as an example of a medium with a significantly higher mineral content. At 20°C, a density was measured that was 27,637 mg/l higher than that of distilled water. Applying the determined correlation results in an estimated mineral content of around 30 g/l – slightly below the actual value of 36–39 g/l. Since sodium chloride is known to be the main component, the specific concentration model of the DLO-M2 could be used for NaCl. The result of 3.88% (w/w) is very close to the expected value.

This series of measurements shows that the DLO-M2 is not only suitable for analyzing drinking and factory water with low mineral content but also delivers reliable results for highly concentrated salt solutions.

Welche Sensoren wurden verwendet?

Sensors that might interest you

Viscosity

Viscosity sensor VLO-M2

Applications that might interest you

Wink of Knowledge: Measuring the density and viscosity of ink in industrial processes

Density and viscosity measurements of ink for inkjet printers using the VLO-M2 viscosity sensor show a high degree of consistency with control measurements in laboratory analyzers over a temperature range of 10 to 65 °C.

Results

The density measurement results are shown in Figure 1. The blue circles and the linear trend curve of these measured values show the measurement results of the laboratory analysis. These were determined using the DSA 5000 M laboratory measuring device from Anton Paar. The squares shown in orange and their trend line show the measured values recorded using the VLO-M2 viscosity sensor. A linear change in density can be seen across the entire temperature range.

Figure 1: Density from 5 to 65 °C

Congruent with the density measurement, Figure 2 shows the viscosity measurements and their trend line, which is represented by a polynomial instead of a linear dependence.

Figure 2: Viscosity from 5 to 65 °C

The viscosity measurement shows that the sensor’s measuring points do not completely match those of the laboratory measuring device. With a measurement deviation of -0.185 mPas, just under ±2%, in viscosity, these results are still well within the sensor’s specification, which is ±[0.2 mPa s + 5% of the measured value]. As mentioned at the beginning, viscosity is a fundamental parameter for printing processes, and viscosity is often adjusted by heating the print head. The VLO-M2 allows both the temperature and viscosity of the ink to be measured in real time, enabling perfect control of the printing process.

These deviations can also be explained by a temperature gradient in the sensor that occurs during heating and cooling in the oven. The parts in direct contact with the warmer (or cooler) ambient air assume a different temperature than the parts and components inside the sensor.

Under real process conditions, where the temperature of the medium and the ambient temperature are stable, a homogeneous temperature can be established throughout the sensor. This means that deviations in viscosity measurement of <2% are plausible.

Procedure

The density and viscosity of an ink for use in industrial printing systems were measured using the VLO-M2 and the DSA 5000 M and SVM 3001 laboratory devices (Anton Paar).

To do this, the VLO-M2 viscosity sensor was completely filled with ink and subjected to a temperature ramp using an oven. Measurement values were recorded continuously. Temperature levels in the range of 5 to 65 °C were set at intervals of 5 °C using the laboratory measuring devices.

These measurements were carried out immediately one after the other to minimize changes in physical properties, for example due to aging effects or moisture ingress.

To ensure comparability of the data, the measurement points recorded in the VLO-M2 viscosity sensor were averaged within a temperature range of ±0.2 °C around the respective level of the laboratory measuring device. This allows the changes in density and viscosity to be graphically represented and estimated.

Conclusion

This test shows that both the density and viscosity of the ink provided for test measurements can be determined very accurately. Compared with the results of the laboratory analysis, a maximum measurement error of ±2% in viscosity and ±0.125% in density can be identified across the entire temperature range measured.

It should be noted that the measurement deviation from the reference can be further reduced by optimizing the measurement setup, making measurement errors in the range of ±1% of viscosity realistic.

With the help of these two parameters, any printing processes where high value is placed on quality and reproducibility can be better monitored and continuously improved.

Which sensors were used?

Sensors that might interest you

Viscosity

Viscosity sensor VLO-M2

Applications that might interest you

Wink of Knowledge: Viscosity measurement by differential pressure and flow rate

This knowledge wink deals with the viscosity determination of media above the measuring range of the dedicated viscosity sensor VLO-M2. Various media were measured in a wide temperature range, whereby viscosities > 400 mPa∙s were achieved. An FLT-M1_i1 Coriolis sensor was used for the flow measurement. The Coriolis measuring principle is ideally suited for this method thanks to the precisely shaped measuring tube over which the pressure drop is measured.

Results

An FLT-M1_i1 was equipped with a pressure sensor at the inlet and a pressure sensor at the outlet of the device (see Figure 1). The measured variables flow rate, inlet pressure, outlet pressure and temperature were recorded for different media and temperatures.

 

Various liquids, from water to high-viscosity ISO 100 hydraulic oil, were pumped through the sensor setup in a temperature-controlled circulation system. Viscosities of approx. 0.5 mPa∙s to approx. 450 mPa∙s were achieved at temperatures between 10 °C and 70 °C.

 

Flow rate and density are included “free of charge” as valuable additional measured variables.

Conclusion

Which sensors were used? 

Sensors that might interest you

Viscosity

Viscosity sensor VLO-M2

Applications that might interest you

More Winks of Knowledge that might interest you

Winkle of Knowledge: Concentration measurement protein

This knowledge wink is dedicated to measuring the concentration of protein in water using the physical parameters of density and viscosity. Commercially available whey protein was used as an example, the concentration of which we were able to determine in the range of ±0.07 %w using a VLO-M2. Our shake is now perfect, and we are also happy to help you optimize your protein solutions!

Results 

Mixtures with different concentrations of commercially available whey protein isolate in water were prepared using a balance. The concentrations were chosen to cover the range close to the manufacturer’s recommendation. The recommendation is to dissolve 25g of powder (which corresponds to approx. 3 level tablespoons) in 300 ml of water. Assuming a density of approx. 1kg/l for water, this corresponds to a concentration of approx. 8.3 %w. Our mixtures ranged accordingly from 4.5 %w to 12.5 %w. The density and viscosity of these mixtures were measured with a VLO-M2 at ambient conditions (approx. 24°C, atmospheric pressure) and yielded the following concentration dependencies (blue dots):

 

 

Both graphs show a very good dependence of the measured values on the protein concentration. For density, the relationship is almost linear, and for viscosity it can be approximated very well as a square (dashed lines). Taking into account the measuring accuracies of the VLO-M2 for the two physical quantities, which are ±0.2 kg/m³ for density and ±0.2 mPas for viscosity, density is clearly the best choice for calculating the concentration under the conditions investigated. However, viscosity as an additional parameter can be very interesting if an additional component such as sugar or alcohol/solvent is added. Even at higher concentrations, the viscosity should then react very strongly to changes in concentration due to the quadratic dependence.

 

The linear fit for the density dependence shows a gradient of 2.74 kg/m³ per weight per cent. With the high measurement accuracy of ±0.2 kg/m³ of the VLO-M2, it would therefore be possible to determine the concentration of whey protein in water with an accuracy of approx. ±0.07 %w (after field adjustment/repeatability even half of this).

 

Of course, with this knowledge we had to directly check the preparation recommendation of 3 level tablespoons in 300 ml of water and ended up at the orange dot in the graphs above. The expected concentration would be 8.3 %w, but the measured density would correspond to a much lower concentration of approx. 5.5 %w. An increase in the amount of protein by approx. 50% would therefore be necessary in our case in order to fulfil the consumption recommendation. As a side note, the water used for the ‘real’ protein shake was slightly cooler (approx. 22°C), which is why the viscosity in this example was also higher. The effect on the water density is relatively small at approx. 0.4 kg/m³ for the 2°C difference and would correspond to approx. 0.15 %w whey protein in this calculation.

Conclusion

Which sensors were used? 

Sensors that might interest you

Viscosity

Viscosity sensor VLO-M2

Applications that might interest you

More Winks of Knowledge that might interest you

Wink of Knowledge: High Density Media – DLO Density Meter for Liquids

Volume 2 | Number 3

What is a wink of knowledge?

Density measurement

DLO density sensor for liquids

The measuring system in submillimetre size enables the compact construction of the sensor. It measures just 80 x 30 x 15 mm and thus fits into even the tightest of spaces. The measured values reach the higher-level system via an RS232 interface and in the ASCII command protocol in the TrueDyne Sensors standard.

Figure 1 – Measurement setup

Results

Figure 2: Measurement results of TrueDyne DLO-M1 density sensors with tetrachloroethylene

Summary

Sensors that might interest you

Viscosity

Viscosity sensor VLO-M2

Applications that might interest you