Correct weighing of the ingredients is essential for reaching the required product quality. If the various raw materials are not dispensed in the correct quantities to begin with, it will never be possible to produce a correct product. For that reason, the dosing of all solid & liquid ingredients into the continuous mixer is constantly monitored. The line stops and gives an alarm signal if a dosage deviates from the set value (the time and % of the deviation are adjustable). But, what is dosing accuracy and how do we measure it?
REPEATABILITY VS. LINEARITY
To fully define feeder accuracy it is important to look at two seperate principles of feeder performance. Only when evaluating a combination of these factors, one can make substantiated statements on dosing accuracy. Lets start at the repeatability factor.
The repeatability measurement is made by taking a series of carefully timed consecutive samples from the discharge stream, weighing them, and then calculating the +/- standard deviation of sample weights expressed as a percentage of the mean value of the samples taken.
Simply said, repeatability is the ability to return to the same position multiple times under identical conditions. However, it is important to understand that the repeatability numbers reveal nothing at all about whether the feeder is delivering, on average, the targeted rate. It only says something about how much on average your weights deviate from the total average of your samples.
Then we have the linearity factor which is evenly important as the repeatability factor – the linearity factor says something about how much your data, within the operating range of the dosing system, on average deviates from your set value. The so called operating range is the range of flow that a dosing system is able to measure with acceptable accuracy. Basically linearity is the percentage of your setpoint. However, only looking at this number does not say sufficient about your dosing accuracy.
For a real-world example, let’s consider a basketball player. If the player is linear, he’ll always get the ball close to the hoop. If his shooting is repeatable, he’ll always shoot to the same location (hopefully, in the basket). The best players are both accurate—hitting the basket—and repeatable—doing it every time.
Lets now go to a reallife example in relation to the continuous dosing of raw materials that will help explain why it is important to look at both factors, linearity and repeatability when talking about dosing accuracy.
Imagine you have a set value of 100 kg/hr and a mean (average) process value of 100.220 kg/hr. In that case your dosing system seems very accurate as it has a linearity percentage of 0.22% (!). However, when looking at the repeatability factor – it might be that the spread of your weights is quite large. If you look at the graph below you can see that eventhough your average value is really close to your set value – there are many moments in time that your system is performing off spec.
In other words, when talking about dosing accuracy – it is crucial to evaluate both your linearity as well as repeatability.
Our experience shows that our and many feeding solutions on the market are nowadays so advanced that they can have a repeatability of 0.5%, at for example 2 sigma (meaning that 95.5% of your weights fall within +/-0.5% bandwith from your mean) and a linearity number of 0.25% meaning that within the operating range of your feeder your weights deviate +/- 0.25% from your set value. This results in a product quality that is extremely consistent and completely independent of human error.
Please note these numbers are based on laboratory conditions and do not account for possible external factors that have to be taken into consideration and act upon when commissioning the production line.
Do you have any questions in relation to dosing accuracy or other processes involved in the continuous production of (food) products? Ask them below.
Continuous regards, Team Sobatech